Paired synovial fluid, plasma and synovial tissue specimens were collected from eight RA and three undifferentiated AC patients. The diagnosis of AC patients stayed unchanged during 1 year of follow-up. The RA patients fulfilled the criteria of the 1987 Criteria of the American College of Rheumatology. The study was approved by the Medical Ethical Committee of the Academical Medical Center of the University of Amsterdam, and informed consent was obtained to participate in the present study. The demographic and clinical data are summarized in Table 1.

Anti-CD4 labeled with phycoerythrin (PE; CLB-T4/2 6D10, IgG₁) and anti-CD66e-PE (CLB-gran/10 IH4Fc, IgG₁) were obtained from the Central Laboratory of the Netherlands Red Cross Blood Transfusion Service (CLB; Amsterdam, The Netherlands), anti-glycophorin A-PE (JC159, IgG₁) was from DakoCytomation (Glostrup, Denmark). Anti-CD8-PE (Leu™-2a, IgG₁), anti-CD14-PE (MφP9, IgG_2b), anti-CD20-PE (L27, IgG₁), anti-CD61-PE (VI-PL2, IgG₁) and IgG₁-PE (X40) were from Becton Dickinson (BD, San Jose, CA, USA), and anti-IgG_2b-PE (MCG2b) was from Immuno Quality Products (Groningen, The Netherlands). IL-6, IL-8 and intracellular adhesion molecule-1 (ICAM-1; Diaclone Research, Besançon, France) and MCP-1, RANTES, VEGF and GM-CSF (BioSource International, Camarillo, CA, USA) were determined by ELISA. IL-1β was obtained from Roche Diagnostics (Mannheim, Germany)

Synovial tissue was collected from an actively inflamed joint by small-needle arthroscopy under local anesthesia with a 2.5 mm biopsy forceps to sample from different areas throughout the knee joint 24. Synovial tissue was placed in Dulbecco's modified Eagle's medium (Life Technologies, Paisley, Renfrewshire, UK) supplemented with 10% FCS, 50 μg/ml streptomycin, 50 IU/ml penicillin and 2 mM L-glutamine and subjected to tissue digestion within 2 hours, as described previously 25. The cells were cultured at 37°C and 5% CO₂. After the second passage, FLS were seeded into 24-well flat-bottomed plates (Costar, Acton, MA) and maintained for 24 hours in culture medium containing 1% FCS.

Immediately before the arthroscopy, we collected synovial fluid (4.5 ml) from the same joint and also venous blood (4.5 ml) into tubes containing 0.5 ml of 3.2% sodium citrate (BD). Immediately after collection, a further 0.5 ml of 3.2% sodium citrate was added to the synovial fluid to prevent clotting. Cells were removed from both blood and synovial fluid by centrifugation for 20 min at 1,550 g and 20°C. For all determinations, aliquots of cell-free plasma and synovial fluid were snap-frozen in liquid nitrogen for at least 15 min and stored at -80°C until use.

For flow-cytometric analysis, cell-free synovial fluid aliquots (250 μl) were thawed on melting ice and centrifuged for 30 min at 17,570 g and 20°C to pellet the microparticles. Supernatant (225 μl) was removed and microparticles were resuspended in 225 μl PBS (154 mM NaCl, 1.4 mM phosphate, pH 7.4), containing 10.9 mM trisodium citrate. After centrifugation for 30 min, supernatant (225 μl) was again removed and microparticles were resuspended in 150 μl of PBS/citrate buffer. For the FLS experiments, microparticles were isolated from 1 ml of synovial fluid by centrifugation for 1 hour at 17,570 g and 20°C. Supernatant (975 μl) was removed and replaced by 975 μl of PBS containing trisodium citrate. Microparticles were resuspended and again pelleted by centrifugation for 1 hour at 17,570 g and 20°C. Again, 975 μl of supernatant was removed and microparticles were resuspended in the remaining 25 μl. This microparticle suspension was added to a final volume of 1 ml of culture medium in which FLS had been maintained for 24 hours. Where indicated, a higher concentration of microparticles was also tested for its ability to activate FLS when sufficient synovial fluid was available. These microparticles, isolated from 3 ml of synovial fluid, were also concentrated into 25 μl of PBS containing trisodium citrate. Microparticle suspensions were each added to FLS cultures from the same donor to mimic the situation in vivo as much as possible.

FLS were quiescent after incubation for 24 hours in medium containing 1% FCS. After 24 hours, this medium (1 ml) was replaced by culture medium containing 1% FCS without any other addition (1 ml; control), or by 975 μl of culture medium plus (1) 25 μl of IL-1β (125 pg/ml final concentration), (2) 25 μl of microparticle suspension or (3) 25 μl of microparticle-free synovial fluid that had been diluted 1:9 in PBS (that is, containing 2.5 μl of the original synovial fluid; this quantity was chosen arbitrarily to correct for both the onefold (unconcentrated) and threefold concentrated microparticle suspensions that, after washing of the microparticles, still contained about 0.7 and 2.1 μl of synovial fluid, respectively). Because individual FLS cultures showed a considerable variation in (mediator) response to the positive control, namely IL-1β, we expressed the response of each FLS culture to microparticles as a percentage of the IL-1β-induced response.

Microparticles were measured by flow cytometry with a method that differed slightly from that used previously 4. In the present study, the microparticles were not washed by centrifugation after being labeled with antibodies because this resulted in a selective loss of microparticle populations. In brief, 5 μl of the microparticle suspension was added to a mixture of PBS (45 μl) containing 2.5 mM CaCl₂ and 5 μl of PE-labeled mAb, and incubated for 15 min in the dark at ambient temperature (20 to 22°C). The following (final concentrations) of mAbs were used: anti-CD4-PE (0.5 μg/ml), anti-CD8-PE (0.25 μg/ml), anti-CD14-PE (0.25 μg/ml), anti-CD20-PE (0.5 μg/ml), anti-CD61-PE (0.5 μg/ml), anti-CD66e-PE (0.25 μg/ml) and anti-glycophorin A-PE (0.25 μg/ml). PE-labeled IgG₁ and IgG_2b (both at 0.5 μg/ml) were used as isotype-specific control antibodies. After incubation, 900 μl of PBS/CaCl₂ was added. Samples were analyzed on a FACSCalibur (BD) and data were analyzed with CellQuest™ Pro software (version 4.02; BD). Both forward scatter and side scatter were set at logarithmic gain. Microparticles were identified by forward scatter, side scatter and binding of cell-specific mAb. The number of microparticles per liter of plasma or synovial fluid was estimated by using the number of events (N) of cell-specific mAb-binding microparticles after correction for control antibody binding: number/liter = N × (150/5) × (955/67) × (10⁶/250). The lower detection limit of the particle count was previously established as 10⁷ microparticles per liter. In this formula, 150 (μl) is the final volume of the washed microparticle suspension, 5 (μl) is the volume of this suspension that is used for each labeling, 955 (μl) is the total volume of the microparticle suspension after labeling before fluorescence-activated cell sorting analysis, 67 (μl) is the average volume of the labeled microparticle suspension that is analyzed by the flow cytometer in 1 min, 10⁶ is the conversion from μl to liter, and 250 (μl) is the original volume of the plasma or synovial fluid sample used for microparticle isolation.

Data were analyzed with GraphPad Prism for Windows, release 3.02 (San Diego, CA, USA). Differences in the concentrations of chemokines, cytokines and VEGF between synovial fluid and plasma as well as in culture supernatants were analyzed with the Wilcoxon signed-rank test. Two-tailed significance levels were considered significant at P < 0.05. All data are presented as medians (range).

Previously, we found no differences between the numbers and cellular origin of microparticles in synovial fluid from RA and AC patients 4. For all cell-specific antigens tested, the microparticle numbers of the three AC patients fell within the range of the RA patients, which is consistent with these earlier observations. The data in Table 2 therefore summarize the microparticle numbers for RA and AC patients together. Most microparticles originated from monocytes (CD14) and granulocytes (CD66e). Microparticles derived from platelets (CD61) and erythrocytes (glycophorin A) were below detection level (less than 10⁷/l) in synovial fluid from all patients, except in one RA patient who had a low but detectable number (1.7 × 10⁷/l) of platelet-derived microparticles. One other RA patient had a relatively high number of erythrocyte-derived microparticles (3.1 × 10⁹/l). Microparticles from CD4⁺ cells were found in six RA patients and all AC patients. Microparticles from CD8⁺ T cells were present in the synovial fluid of five RA patients and one AC patient. Microparticles from B cells were found in two RA patients only.

FLS were quiescent after incubation for 24 hours in medium containing 1% FCS. The concentrations of all markers studied in the FLS culture supernatants are summarized in Table 3. In comparison with the control (unstimulated), IL-1β significantly increased the levels of all mediators tested, whereas the addition of microparticle-free synovial fluid affected especially the soluble ICAM-1 (sICAM-1) levels. This increase was due to its presence in the synovial fluid itself. Addition of microparticles to FLS significantly increased the levels of MCP-1 (P = 0.010), sICAM-1 (P = 0.010), IL-8 (P = 0.008), IL-6 (P = 0.042), VEGF (P = 0.001) and RANTES (P = 0.031). In contrast, the concentrations of GM-CSF decreased (P = 0.002).

In six patients (three RA and three AC patients), we also tested a threefold higher (final) concentration of synovial microparticles. In comparison with the 'onefold' concentration, levels of sICAM-1 (P = 0.031), IL-8 (P = 0.031) and IL-6 (P = 0.031) increased further and GM-CSF (P = 0.016) decreased further (Table 3). Levels of MCP-1 (P = 0.156), VEGF (P = 0.078) and RANTES (P = 0.062) also tended to increase further, but these differences did not reach statistical significance.

Because individual microparticle suspensions were tested in (autologous) FLS cultures and considerable differences were observed in the responsiveness of these individual cell cultures, the individual responses of FLS cultures are also shown (Fig. 1). The response is expressed as either an increase or a decrease relative to the control, namely the 24-hour incubation of FLS with the microparticle-free synovial fluid. Although variation between FLS cultures is apparent, the individual data substantiate the conclusions above as based on group analysis.

For comparison, the concentrations of the various mediators were also determined in both synovial fluid and plasma from RA and AC patients. Because only 2 values (of 36) of the AC patients fell outside the RA range, namely MCP-1 in synovial fluid and sICAM-1 in plasma from the same AC patient, all data are summarized in Table 4. In comparison with plasma, levels of MCP-1 (P = 0.008), IL-6 (P = 0.002), IL-8 (P = 0.002) and VEGF (P = 0.002) were elevated in synovial fluid, those of RANTES and ICAM-1 were decreased (P = 0.001 and P = 0.006, respectively), and GM-CSF concentrations were similar (P = 0.125). Figure 2 shows that both the total number of microparticles (Fig. 2a; r = 0.91; P < 0.0001) and the numbers of granulocyte-derived microparticles (Fig. 2b; r = 0.89, P < 0.0001) were correlated with the IL-8 concentrations, whereas the numbers of monocyte-derived microparticles were not (Fig. 2c; r = 0.04; P = 0.89). In addition, concentrations of MCP-1 were correlated with total numbers of microparticles (r = 0.81, P < 0.0001) and numbers of granulocyte-derived microparticles (r = 0.93, P < 0.0001), but again not with the numbers of monocyte-derived microparticles (r = 0.06; P = 0.82; data not shown). No other correlations were found between microparticle numbers and concentrations of mediators.