Cytokines in osteoarthritis: mediators or markers of joint destruction?

OBJECTIVE: The integrity of articular cartilage is maintained by the balance between cytokine-driven anabolic and catabolic processes. Unregulated or excess influences of these molecules are thought to play a part in the pathophysiology of many joint diseases. However, the role of cytokines in osteoarthritis (OA) is not well established. Our aims are twofold: firstly to consider the evidence for the contribution of cytokines to the pathophysiology of OA and secondly to evaluate their potential as markers of disease activity in OA. METHODS: Cytokine homeostasis, the role of catabolic and anabolic cytokines in maintaining cartilage integrity, and the contribution of such cytokines to destructive processes in OA were examined. Consideration was given to the interrelationship between cartilage, bone, and synovium in OA; metabolites produced by such structures were compared with cytokines as indicators of disease activity. RESULTS: The evidence reviewed suggests that interleukin-1 (IL-1) and the less potent tumor necrosis factor alpha (TNF alpha) are mediators of joint damage in OA. The cytokines interleukin-6 (IL-6) and leukemia inhibitory factor (LIF) were implicted in both destructive and protective mechanisms, suggesting a dual role. Metabolites of the different components of the joint provided a better measure of disease activity than cytokines. CONCLUSIONS: Experimental evidence is emerging that catabolic cytokines are mediators of joint damage in OA, although their usefulness as markers of disease activity is limited because of the need to monitor a wide range of ligands and their inhibitors simultaneously. In contrast, metabolites released from cells within bone, synovium, and cartilage related to disease activity and provided prognostic information.

Mediators of joint swelling and damage in rheumatoid arthritis and pristane induced arthritis.

Joint swelling and tenderness in rheumatoid arthritis (RA) probably result from IgG aggregates activating complement with the consequent attraction of polymorphonuclear leucocytes (PMNs) and the liberation of their granule enzymes such as kininogenases. By contrast IL-1 and TNF are the major stimulants of cartilage and bone loss although other agents contribute. The fundamental drive for the production of these mediators is unknown but a role for heat shock proteins is suggested from work on pristane induced arthritis.

The effects of pressure on chondrocyte tumour necrosis factor receptor expression.

This work was performed to determine whether one aspect of load, pressure, could alter tumour necrosis factor (TNF) receptor type I (RI) expression on chondrocytes. Encapsulated tsT/AC62, osteoarthritic (OA) or non-arthritic (NA) chondrocytes were centrifuged at speeds representing 5 or 20 MPa, incubated for specific periods, released from alginate and TNFRI and II (TNFRII) expression determined by flow cytometry. Significant (p<0.05, n=4) changes in tsT/AC62 chondrocyte TNFRI expression were apparent 24 hours after application of 20 MPa. Five or 20 MPa increased OA chondrocyte TNFRI expression; chondrocytes from some OA patients were markedly sensitive to 20 MPa. NA chondrocyte TNFRI expression usually decreased in response to 5 and 20 MPa. Significant pressure-induced differences in TNFRI expression between NA and OA groups were apparent at 5, but not 20 MPa. Pressure did not significantly alter TNRFII expression on tsT/AC62, NA or OA chondrocytes. These results suggest a mechanism whereby sensitivity of chondrocytes to the effects of TNFalpha may be increased, in susceptible individuals, in regions of the joint that experience peak loading.

Interleukin 1 alpha and beta production by cells isolated from membranes around aseptically loose total joint replacements.

Aseptic loosening of joint prostheses is accompanied by local osteolysis. To determine whether local production of interleukin 1 might contribute to such lysis, the number of interleukin 1 secreting cells in the pseudosynovial membrane surrounding prostheses was measured. Interleukin 1 alpha and beta secreting cells were identified by ELISPOT, a sensitive cytokine secreting assay. The proportion of interleukin 1 beta secreting cells in pseudosynovial membrane was comparable with the proportion occurring in the synovium of patients with rheumatoid arthritis and higher than that in normal subjects and patients with osteoarthritis. The proportion of interleukin 1 alpha producing cells was higher in pseudosynovial membrane than in diseased synovium. Overall, higher numbers of interleukin 1 beta than interleukin 1 alpha secreting cells were detected. A correlation was observed between the number of cells in pseudosynovial membrane producing interleukin 1 beta and those producing interleukin 1 alpha. When divided into area of origin, tissue samples from the femoral area contained a higher proportion of interleukin 1 beta producing cells than tissue in the acetabular or capsular regions, though due to variance within each group this difference did not reach significance.

IL-1 secreting cell assay and its application to cells from patients with rheumatoid arthritis.

The cells within a population that were secreting interleukin-1 (IL-1) were enumerated and visualized by an ELISA-SPOT assay. Initial experiments designed to validate the assay revealed that the number of IL-1 beta spot forming cells was increased by exposing normal blood monocytes to LPS and that spot formation was prevented by incubating the cells with cycloheximide. Normal blood polymorphonuclear leucocytes (PMNs) produced IL-1 alpha and IL-1 beta in response to recombinant granulocyte monocyte colony stimulating factor (rhGMCSF) but not to cytochalasin B, calcium ionophore or LPS. Monocytes and PMN were isolated from the synovial fluid (SF) and blood of patients with rheumatoid arthritis (RA) and the ability of these cells to secrete IL-1 alpha and IL-1 beta compared. A higher proportion of SF derived monocytes were found to secrete IL-1 beta spontaneously compared to the corresponding blood cells. IL-1 alpha secreting monocytes were not detected although high numbers of IL-1 alpha secreting cells were found among cells isolated from rheumatoid synovium. By contrast SF PMNs did not produce IL-1 alpha or IL-1 beta whereas blood PMNs from some (3/8) RA patients produced IL-1 alpha and/or IL-1 beta. It is considered that the IL-1 ELISA-SPOT is a highly sensitive technique for detecting IL-1 secreting cells.

Osteoarthritic synovial fluid and synovium supernatants up-regulate tumor necrosis factor receptors on human articular chondrocytes.

OBJECTIVE: To determine whether the up-regulation of chondrocyte tumor necrosis factor receptor (TNF-R) expression in osteoarthritis (OA) is due to molecules released within the OA knee joint. DESIGN: Non-arthritic (NA) human articular chondrocytes were incubated with normal serum, OA synovial fluid, or supernatants from either cultured NA or OA synovium, and TNF-R expression measured by flow cytometry. RESULTS: OA synovial fluid, but not normal serum, significantly up-regulated the proportion of chondrocytes expressing p55 TNF-R as well as the number of p55 TNF-R/chondrocyte. Similarly, supernatants from OA, but not NA, synovia significantly up-regulated chondrocyte p55 TNF-R expression. Chondrocyte p75 TNF-R expression was also significantly increased by some of the OA supernatants but not others, and overall no significant increase was seen. OA synovium supernatants contained higher concentrations of interleukin-1 beta (IL-1 beta) and interleukin-6 (IL-6) than NA synovium supernatants and neutralizing antibodies to these cytokines either partially or totally abrogated the ability of the OA supernatants to increase chondrocyte p55 TNF-R expression. Finally, various concentrations of recombinant human (rh)IL-1 beta and rhIL-6 up-regulated chondrocyte p55 TNF-R expression. CONCLUSION: These results suggest that IL-1 and IL-6 produced by OA synovium contribute to the progression of the disease by rendering chondrocytes more susceptible to stimulation by catabolic cytokines.

Chondrocyte tumor necrosis factor receptors and focal loss of cartilage in osteoarthritis.

OBJECTIVE: Osteoarthritis (OA) is characterized by focal loss of cartilage. Here we show for the first time that tumor necrosis factor (TNF) alpha can act on cartilage but only at specific sites where chondrocyte TNF alpha-receptor (R) expression is high. DESIGN: Cartilage explants from specified sites in the knee joints of both OA patients and non-arthritic (NA) subjects were cultured with and without TNF alpha for 14 days and cumulative glycosaminoglycan (GAG) release into the supernatant measured. p55 and p75 TNF-R expression was measured by flow cytometry on chondrocytes isolated from the same sites. RESULTS: Cartilage explants from different sites in knee joints from both OA patients and NA subjects varied in their susceptibility to TNF alpha. Overall, the proportion of samples that responded to TNF alpha was higher in cartilage taken from OA joints than cartilage from NA subjects. Variations in p55 and p75 TNF-R expression were found between chondrocytes from different sites. p55, but not p75 TNF-R, expression on chondrocytes was closely related to the susceptibility of explants from the same site to TNF alpha-induced GAG loss. CONCLUSION: It is considered that focal loss of cartilage will occur at sites where chondrocyte p55 TNF-R expression is high, if sufficient TNF alpha is present, and that these results identify a mechanism by which cytokine-mediated focal loss of cartilage may occur.

Susceptibility to physiological concentrations of IL-1beta varies in cartilage at different anatomical locations on human osteoarthritic knee joints.

OBJECTIVES: To determine whether cartilage biopsies from specific regions of osteoarthritic knee joints differ in susceptibility to the degradative effects of the amounts of interleukin-1 beta (IL-1 beta; 1-10 pg/ml) found in osteoarthritic joints. To establish whether biopsies are sensitive to the effects of either IL-1 beta or TNFalpha or both catabolic cytokines. METHODS: Cartilage from specified regions of 22 osteoarthritic knee joints was examined. Biopsies were incubated for 14 days without or with IL-1 beta or TNFalpha at physiological concentrations and GAG release into supernatants assessed. RESULTS: Variation was observed in susceptibility to the effects of 1-10 pg/ml IL-1 beta in biopsies from different sites within the same joints and the same site in different joints. The number of regions responding to the cytokine increased significantly (P< 0.0063, Chi square test) with concentration: only 10% (2/21) of all regions tested were susceptible to the effects of 1 pg/ml IL-1 beta, whereas 45% (9/20) were susceptible to the effects of 5 pg/ml and 56% (10/18) to the effects of 10 pg/ml IL-1 beta. Significantly fewer regions (4%, 2/47) responded to both IL-1 beta and TNFalpha (P< 0.047, Chi square test); biopsies from some patients responded to neither cytokine. CONCLUSIONS: IL-1 beta, at the low concentrations detected in joints, can degrade cartilage from susceptible locations. Susceptibility of some cartilage biopsies to the effects of either IL-1 beta and TNFalpha, but not both, suggests the signalling receptors for the two major catabolic cytokines are not usually expressed concurrently. The fact that some biopsies respond to neither cytokine suggests that in some patients the local concentration of inhibitors may be high or that other catabolic stimuli predominate. These results could have important implications for pharmacological intervention strategies.

Tumor necrosis factor-alpha receptor expression on chondrocytes isolated from human articular cartilage.

OBJECTIVE: To determine whether the previously observed enhanced susceptibility of osteoarthritic (OA) cartilage to tumor necrosis factor-alpha (TNF-alpha) induced degradation could be due to differences in receptor expression. METHODS: Using specific monoclonal antibodies flow cytometry was used to identify and quantify 2 TNF-alpha receptors on the surface of chondrocytes isolated from human articular cartilage. RESULTS: The proportion of chondrocytes expressing p55 TNF-alpha receptor was significantly higher in populations from OA cartilage (p < 0.01) when compared with similar populations from non-arthritic cartilage. The number of p55 receptors/chondrocyte was also significantly higher (p < 0.01) in OA. A significant correlation (r = 0.72, p < 0.18) was found in OA between the numbers of chondrocytes expressing p75 receptor and the number of receptors/chondrocyte. CONCLUSION: p55 TNF-alpha receptor expression is significantly increased on OA chondrocytes ex vivo. Enhanced expression of p55, purported to be the biologically active receptor, could contribute to OA cartilage degradation.

Increased spontaneous production of interleukin-1 together with inhibitory activity in systemic sclerosis.

1. Increased spontaneous production of interleukin-1, measured as lymphocyte-activating factor activity, was seen in unstimulated monocytes from systemic sclerosis patients. 2. Inhibitory activity to interleukin-1 was seen in both normal and patient monocyte supernatants. 3. Inhibitory activity was significantly higher in unstimulated and stimulated monocyte supernatants from systemic sclerosis patients. 4. The net effect was an apparent decrease in lymphocyte-activating factor activity in the monocyte supernatants from systemic sclerosis patients. 5. These findings suggest a possible mechanism by which collagen deposition could be enhanced, thereby giving rise to the extensive fibrosis characteristic of systemic sclerosis.

Alteration of cartilage metabolism by cells from osteoarthritic bone.

OBJECTIVE: To determine whether bone cells alter cartilage metabolism. METHODS: Bone cell cultures were established using explants obtained from the hip and knee joints of 9 patients with osteoarthritis (OA) and 6 subjects without arthritis (nonarthritic [NA]). NA human cartilage biopsy samples were incubated in the presence or absence of bone-derived cells, and the effects on glycosaminoglycan (GAG) release from cartilage were measured. RESULTS: Bone cell cultures secreted osteocalcin (OC) and did not contain cells expressing leukocyte common antigen. None of the 8 cultures established from NA bone, compared with 17 of 32 from OA bone, significantly altered GAG release from cartilage (P = 0.006). In knees with medial joint damage, 38% of the cultures derived from the medial side of the joint increased GAG release from cartilage. In contrast, 77% of the cultures derived from the lateral side of the joint had an effect on GAG, with 38% increasing and 38% decreasing GAG release. Seven cytokines were measured in OA bone cell supernatants. No significant difference was apparent in the concentration of any one cytokine when supernatants were compared according to their effects on GAG release. CONCLUSION: Bone cells from OA patients can influence cartilage metabolism. This might explain why increased subchondral bone activity can predict cartilage loss.

Synovial fluid concentration of five different cytokines in rheumatic diseases.

Interleukin-1 beta, interleukin-2, tumour necrosis factor alpha, and the interferons, alfa and gamma, were measured concurrently in synovial fluid samples from 68 patients with rheumatic diseases. Mean interleukin-1 beta concentrations (130.3 (SD 22) pg/ml) were higher in synovial fluids from patients with rheumatoid arthritis (RA) than in those from patients with osteoarthritis (27.8(4.5)pg/ml), while measurements in synovial fluids from patients with seronegative spondarthritis were intermediate (72.7 (32) pg/ml). Interleukin-2 and tumour necrosis factor alpha concentrations were lower in the inflammatory arthropathies (RA: 4.5 (0.6) U/ml, 0.39 (0.04) ng/ml; seronegative spondarthritis: 3.1 (0.3) U/ml, 0.33 (0.03) ng/ml respectively) than those in patients with osteoarthritis (5.2 (0.6) U/ml; 0.05 (0.04) ng/ml). Interleukin-2 and tumour necrosis factor alpha concentrations correlated in all groups (r = 0.7), as did the interferons alfa and gamma (r = 0.7). There was no relation between interleukin-1 beta and either interleukin-2 or tumour necrosis factor alpha, or between the interferons and any other cytokine. Several distinct cytokine patterns were noted. Synovial fluids from two non-arthritic subjects were also examined: interleukin-1 beta concentrations were low, but concentrations of the other cytokines were higher than those seen in most arthritic fluids.

Tumor necrosis factor alpha can contribute to focal loss of cartilage in osteoarthritis.

UNLABELLED: OBJECTIVE; To evaluate the potential for tumor necrosis factor alpha (TNFalpha)-induced focal loss of cartilage in osteoarthritic (OA) knee joints. DESIGN: Fresh cartilage from specified regions of OA joints was immunostained for TNF-receptor (R) bearing chondrocytes. Cartilage explants from the same regions were cultured with or without small amounts of TNFalpha and cumulative GAG release into supernatants measured. Concentrations of TNFalpha, p55 and p75 soluble (s) TNF-R in supernatants from cultured OA and non-arthritic (NA) synovium were measured by ELISA. RESULTS: TNF-R bearing chondrocytes were identified in OA cartilage; more specimens contained p55 TNF-R- than p75 TNF-R-bearing chondrocytes and differences in TNF-R distribution were apparent in cartilage from different regions of the same knees. TNFalpha at 5, 1, 0.5 and 0.25 ng/ml (but not 0.1 ng/ml) significantly increased glycosaminoglycans (GAG) release from cartilage explants in a dose-dependent manner. Variation in susceptibility to TNFalpha was observed in explants from different sites. TNFalpha and p75 sTNF-R, but not p55 sTNF-R, concentrations were significantly higher in OA, as compared with NA, supernatants. A significant correlation between TNFalpha and p75 sTNF-R measurements was apparent only in NA supernatants. CONCLUSIONS: Variations in chondrocyte TNF-R expression occur in OA cartilage in vivo. TNFalpha at concentrations produced by OA synovium in vitro, can degrade cartilage matrix. In most OA supernatants sTNF-R concentrations were insufficient to abrogate the effects of TNFalpha. Thus conditions exist in some OA knees for TNFalpha to contribute to focal loss of cartilage.