Platelet-activating activity in synovial fluids of patients with rheumatoid arthritis juvenile rheumatoid arthritis gout and noninflammatory arthropathies.код для вставкиСкачать
800 PLATELET-ACTIVATING ACTIVITY IN SYNOVIAL FLUIDS OF PATIENTS WITH RHEUMATOID ARTHRITIS, JUVENILE RHEUMATOID ARTHRITIS, GOUT, AND NONINFLAMMATORY ARTHROPATHIES CHRISTINE SHAPLEIGH, FRANK H. VALONE, PETER H. SCHUR, EDWARD J. GOETZL, and K. FRANK AUSTEN The predominant activating activity in rheumatoid synovial fluid for human platelets, as assessed by the release of I4C serotonin, was previously isolated and shown to be a complex of IgG of molecular weight 450,000. Thus, a pooled 5 ml concentrate of the 450,000 molecular weight fraction obtained by Sepharose 6B gel filtration of 1 ml of synovial fluid was used to compare the platelet-activating activity (PAA) of fluids from 28 patients with inflammatory and noninflammatory arthropathies. One unit of PAA was arbitrarily defined as that volume of the 5 ml concentrate required to obtain release equal to that obtained with a standard dose of thrombin, and total units per milliliter of synovial fluid were calculated as the reciprocal of the volume of concentrate in milliliters containing one unit multiplied by five. A platelet-activating activity of >766 units/ml of synovial fluid was arbitrarily defined as elevated, since this value represents two standard deviations above the mean for the noninflammatory synovial fluids. Each of 7 noninflammatory fluids (range <50 to 695) and 3 of 4 gouty fluids (range <50 to 1,250) fell within the normal From the Departments of Medicine, Harvard Medical. School, and the Robert B. Brigham Hospital Division of the Affiliated Hospitals Center Inc., Boston, Massachusetts 021 15. Supported in part by Grants AI-07167, AI-07722, AM-I 1414, AM-05577, and RR-05669 from the National Institutes of Health, and by a grant from The New England Peabody Home for Crippled Children. Christine Shapleigh, MD: Postdoctoral Trainee, National Institutes of Health; Frank H. Valone, MD: Research Associate of the Howard Hughes Medical Institute; Peter H. Schur, MD, Edward J. Goetzl, MD: Director of the Howard Hughes Institute Laboratories for the Study of Immunological Diseases; K. Frank Austen, MD. Address reprint requests to Dr. K. Frank Austen, Room 604, Seeley G. Mudd Building, 250 Longwood Avenue, Boston, Massachusetts 021 15. Submitted for publication January 11, 1980 accepted in revised form March 20, 1980. Arthritis and Rheumatism, Vol. 23, No. 7 (July 1980) range, whereas 9 of 13 adult rheumatoid fluids (range <50 to 11,880), and 4 of 4 JRA fluids (range 1,299 to 5,100) had elevated levels of PAA per ml of synovial fluid. The finding that the purified synovial fluid 450,000 molecular weight complex of IgG was inhibited in its platelet-activating capacity by purified monoclonal rheumatoid factor suggests that a role for IgM rheumatoid factor could be to block IgG complex Fc receptor-dependent proinflammatory effects on platelets, and perhaps other cells. Intermediate molecular weight complexes of IgG present in the serum of patients with rheumatoid arthritis (1) have been shown to contain IgG rheumatoid factors (2) which preferentially associate with each other to the exclusion of IgG molecules of other specificities (3). Certain rheumatoid synovial fluids contain IgG complexes which were initially considered to resemble IgG aggregated by physicochemical procedures (4) but were subsequently shown to be immune complexes composed of IgG rheumatoid factors (5). Both experimentally formed IgG-containing immune complexes and aggregated IgG induce platelet aggregation and secretion of serotonin by interaction with the Fc receptor (6,7). Although platelet aggregation has been used as an assay of immune complexes in serum (8), it has not been applied to synovial fluid. Platelet-activating activity (PAA), as assessed by the release of I4Cserotonin in the absence of lactic acid dehydrogenase from human platelets, has been identified in rheumatoid synovial fluid and purified by cation exchange chromatography, gel filtration, and preparative isoelectric focusing to yield a 450,000 molecular weight principle termed purified synovial fluid platelet-activating activity (SF-PAA); SF-PAA PLATELET-ACTIVATING ACTIVITY contained only IgG by immunochemical analysis (9). T h e level of platelet-activating activity in a 450,000 molecular weight fraction of synovial fluid has been found to be elevated in adult and juvenile rheumatoid arthritis but not in gout or noninflammatory arthropathies. MATERIALS AND METHODS 5-[2-’4C]hydr~xytryptamine binoxalate (14C serotonin, specific activity 47 mCi/mmol), Aquasol Universal LSC cocktail (New England Nuclear Corp., Boston, Massachusetts), human thrombin (Fibrinidex, Ortho Diagnostics Inc., Raritan, New Jersey), five times recrystallized human serum albumin (Miles Laboratories, Inc., Kankakee, Illinois), and Sepharose 6B (Pharmacia Fine Chemicals, Uppsala, Sweden) were obtained as noted. Platelet suspensions were prepared with siliconized glassware (Pro-Sil-28, Fisher Scientific Co., Pittsburgh, Pennsylvania) or with plasticware (Walter Sarstedt Inc., Princeton, New Jersey). Protein concentration was determined by optical absorbency at 280 nm on a Gilford spectrophotometer (Oberlin, Ohio). Protein concentration of synovial fluid was assessed by the method of Lowry et a1 (lo), synovial fluid rheumatoid factor titers were determined by nephelometry (1 1), and serum rheumatoid factor titers by latex fixation (12) or nephelometry. Clq binding (13), inhibition of antibody-dependent cellular cytotoxicity (ADCC) (14,15), and the complement components Clq, C4, C3 (16), and B (17) were measured as described. Purified monoclonal IgM rheumatoid factor (mRF) containing 3.26 mg/ml IgM and less than 0.031 mg/ml IgG was obtained from Dr. Vincent Agnello (18). SF-PAA was isolated as described and stored at a concentration of IgG of 1.4 mg/ml (9). Patient population. Synovial fluids were obtained from 28 patients (Table 1). The thirteen patients with adult rheumatoid arthritis satisfied the American Rheumatism Association criteria for definite or classic rheumatoid arthritis (RA) (19). Eleven were seropositive with rheumatoid factor titers greater than 1:64, and 2 were seronegative. The diagnosis of juvenile rheumatoid arthritis (JRA) was based on the criteria established by the JRA subcommittee of the American Rheumatism Association (20), and subtypes were classified by observations made in the first 6 months of disease. Three of the JRA patients were females with oligoarticular disease without iridocyclitis and with negative serum titers for rheumatoid factor and antinuclear antibodies (ANA). The male JRA patient had oligoarticular inflammatory arthritis, psoriatic skin lesions present from birth, intermittent diarrhea, a negative serum rheumatoid factor, a positive ANA titer of 1 :40, and an HLA-B27 tissue type. The 4 patients with gout had negative birefringent crystals in the synovial fluid; 1 patient had tophaceous disease. Of 7 patients with noninflammatory arthropathies, 3 had osteoarthritis, diagnosed by the radiologic criteria of Kellegren (21), 3 had mechanical derangements, and one had a traumatic arthropathy. Processing of synovial fluid. Synovial fluid was obtained during diagnostic or therapeutic arthrocentesis and was immediately placed on ice. The fluid was transferred to 14 ml plastic tubes, and the cells and other debris were sedimented 80 1 Table 1. Patient characteristics Diagnosis Adult rheumatoid arthritis Juvenile rheumatoid arthritis Gout Noninflammatory arthropathies Number of patients M F Range Mean 13 6 I 26-81 56 4 4 1 3 0 9-28 16 4 35-60 49 7 5 2 30-80 57 Sex Age (years) by centrifugation at lOOOg at room temperature for 10 minutes. Samples of each supernatant were analyzed for PAA immediately or after storage at 4°C for 1 to 5 days; additional samples were frozen immediately at -70°C and assessed later for C l q binding, inhibition of ADCC, rheumatoid factor titer, and total protein concentration. Two milliliters of each unfrozen synovial fluid supernatant were dialyzed against one change of 1 liter of 0.01M Tris buffer, pH 7.4, containing 0.1M NaCl at 4°C overnight, and any residual debris was removed by centrifugation at 7000g for 2 minutes. One milliliter of supernatant was then applied at room temperature to an 80 x 1.8 cm column containing Sepharose 6B equilibrated in the same buffer. Fractions of 3.8 ml(2% of bed volume) were collected at a flow rate of 15 ml per hour, and 1.0 ml portions of every other fraction were dialyzed in separate bags against one change of 4 liters of 0.01M phosphate bbffer, pH 7.0, containing 0.1M NaCl at 4°C overnight; replicate 250 p1 samples of each portion were assayed immediately or after 1 to 14 days of storage at 4°C for their capacity to release I4C serotonin from platelets. Fractions from Sepharose 6B filtering with an approximate molecular weight of 450,000 and yielding more than 10% net I4C serotonin release were pooled along with the adjacent fractions and concentrated to 5 ml at 4°C with an Amicon ultrafiltration apparatus (Amicon Corp., Lexington, Massachusetts). If less than 10% net 14C serotonin release was detected, all fractions from the 350,000-550,000 molecular weight region were pooled and concentrated. The 5 ml concentrate was dialyzed overnight at 4°C against one change of I liter of 0.01M phosphate buffer, pH 7.0, containing 0.1M NaC1, stored at 4°C for 1 to 14 days, and assayed in a doseresponse fashion for I4Cserotonin release. Platelet serotonin release assay. Fifty to 100 milliliters of blood were obtained by venipuncture with sterile plastic syringes from normal adult volunteers who had not taken aspirin-containing compounds for the previous 7 days. The samples were aliquoted by 5 ml portions into 14 ml plastic tubes containing 1 ml of citrate anticoagulant (0.076M citric acid, 0. ISM sodium citrate, pH 5.2). The erythrocytes and leukocytes were sedimented by centrifugation at 170g for 15 minutes at room temperature. The platelet-rich plasma supernatants from each tube were aspirated with siliconized Pasteur pipettes, pooled into a 50 ml plastic tube, and mixed with a final concentration of 0.4 x 10-6M I4C serotonin. After incubation at 37°C for 30 minutes, the labeled platelets were sedimented at lOOOg for 10 minutes, and the cell pellet was washed 3 times with 0.01M phosphate buffer, pH 6.5, containing 0.1M NaCl, 0.1 1M glucose, and 20% (v/v) citrate anti- SHAPLEIGH ET AL 802 coagulant. The platelets were suspended in 0.01M phosphate buffer, pH 7.0, containing 0.1M NaC1, 0.1 1M glucose, 1.8 mM CaCl,, and 0.1% (w/v) human serum albumin to yield a final cell concentration of 2-3 x lox platelets/ml. Two hundred fifty microliters of radiolabeled platelets were added to 3 ml plastic tubes containing 250 p1 of either buffer or test sample in buffer, and the mixture was incubated at 37°C for 30 minutes with gentle agitation. The platelets were sedimented at lOOOg for 10 minutes at room temperature; the supernatant was mixed with 10 ml of Aquasol in 20 ml liquid scintillation vials, and radioactivity was determined in a Searle Mark IV Liquid Scintillation Counter. The total 14C serotonin in 250 pl of washed radiolabeled platelets was determined by adding the sample directly to counting vials containing Aquasol. Results were expressed as the net percent release of 14Cserotonin according to the following formula: Net % serotonin released = (“C serotonin released) - (14Cserotonin leak) x 100 (Total I4C serotonin) - (I4Cserotonin leak) In 74 experiments the spontaneous leak of 14C serotonin by platelet suspensions incubated with buffer ranged from 0.5% to 16.6%, with a mean of 6.6%. Statistical analysis was performed with a one-sided Wilcoxon signed rank test and Spearman’s rank correlation coefficient. RESULTS In preliminary studies crude rheumatoid synovial fluid stimulated release of platelet 14Cserotonin, but a dose-response relationship was observed only after sequential chromatographic purification of the 450,000 molecular weight principle containing IgG, termed SFPAA (9). Thus, a clinical study was initiated with a 450,000 molecular weight fraction of synovial fluid obtained by Sepharose 6B gel filtration, as depicted in Figure 1 for 1 ml of synovial fluid from a patient with rheumatoid arthritis. The fractions containing the peak of platelet-activating activity (PAA) of approximately 450,000 molecular weight, as defined by the position of the ferritin marker, were pooled, concentrated to 5 ml and assayed in a dose-response fashion for platelet I4C serotonin release. Volumes of 100, 10, 2, 1, 0.5, and 0.1 p1 gave a net percent I4Cserotonin release of 52%, 63%, 54%, 42%, 9%, and 0%, respectively. Release was doserelated, reaching a plateau for most synovial fluids, the exceptions generally being synovial fluid with insignificant concentrations of PAA. Maximal 14C serotonin releases (mean -C SD) for fluids from patients with adult rheumatoid arthritis, juvenile rheumatoid arthritis, gout, and noninflammatory arthropathies were 53 k 12%, 59 k 12%, 45 f lo%, and 47 f lo%, respectively. The additional peak of PAA (Figure 1) in the limit vol- Peak of Approx, m MW 450 OOO Figure 1. Platelet-activating activity in eluant fractions from gel filtration on Sepharose 6B after application of l ml of synovial fluid from a patient with rheumatoid arthritis. The column was calibrated with blue dextran (exclusion volume), ferritin (molecular weight 450,000),and B12 (molecular weight 1355). ume was variably present in all patient groups, whereas that at the void volume was present in 8 adult rheumatoid synovial fluids only. These two additional peaks consistently contained less than one-tenth the PAA of the 450,000 molecular weight peak as measured by the dose-response release of I4Cserotonin and are not considered further. To minimize day-to-day variations in platelet sensitivity, a thrombin standard was analyzed in parallel with test samples in a final concentration from 0.01 to 0.40 units/ml which gave a dose-related net percent 14Cserotonin release from 5% to 95%. In 10 consecutive experiments, net percent serotonin release by 0.08 units/ml thrombin was 44 11% (mean k SD), and one unit of synovial fluid PAA was defined as that volume required to obtain a net percent I4C serotonin release equal to that generated by 0.08 units/ml thrombin in that assay. Total units of PAA for 1 ml of synovial fluid were expressed as the total number of units present in the 5 ml concentrate of the 450,000 molecular weight peak and were calculated as the reciprocal of the sample volume containing one unit, expressed in milliliters, multiplied by 5. If all volumes of a sample (up to 100 pl) generated less serotonin release than 0.08 units/ml throdin, units of PAA were expressed as 4 0 . The stability of PAA in crude synovial fluid was assessed by gel filtration before and after storage for 4-6 months at -7OOC. Units of PAA before and after freezing were 11,880 and 10,000, respectively, for one rheumatoid synovial fluid, 65 and 4 0 for a gouty fluid, and 695 and 4 0 for a noninflammatory fluid. Assessment of the reproducibility of PAA was determined for 4 fluids in 3 PLATELET-ACTIVATING ACTIVITY 803 80 60 40 20 0 VOLUM€ O f 5 m l CONCENW'ATE ON Figure 2. Dose response of I4C serotonin release by incremental volumes of the 5 ml concentrate of the 450,000 molecular weight fraction of 1 ml of synovial fluid from patients with seropositive rheumatoid arthritis, gout, and noninflammatory (traumatic) arthritis, each assayed on 3 separate occasions (U, M, A-A). The corresponding release by 0.08 units/ml thrombin on each of these occasions was 30% (O), 34% (0), and 58% (A). experiments over a 6-week period using samples of the 5 ml concentrate which had been stored at 4°C. Units of PAA ranged from 10,640 to 12,500 with a mean of 11,880 and from 5,555 to 10,000 with a mean of 7,350 for 2 rheumatoid fluids, from t50 to 103 with a mean of 67 for one gouty fluid, and from 4 0 to 565 with a mean of 255 for one noninflammatory fluid (Figure 2). Platelet-activating activity, expressed as units of PAA per ml of synovial fluid, is depicted in Figure 3. The mean PAA for the 7 noninflammatory fluids was 264 +. 251 units (mean k SD) per ml and, therefore, fluids were designated as containing elevated PAA if they contained greater than 766 units, which equaled 2 standard deviations above the mean for the noninflammatory fluids. Each of the 7 noninflammatory and 3 of 4 gouty synovial fluids contained less than 766 units of PAA. Nine of the 13 adult rheumatoid synovial fluids had elevated levels of PAA with units ranging from 2,630-1 1,880, while 4 had less than 766 units of PAA. In patients with adult rheumatoid arthritis, units of PAA correlated inversely with synovial fluid C3 protein (rs = -0.74, n = 12, P < 0.01) (Table 2) but not with C l q (rs = -0.13, n = 12, P > O.lO), C4 (r$= -0.39, n = 12, P > 0.10), or B (r. = -0.18, n = 12, P > 0.10) proteins, and there was no apparent relationship between units of PAA and synovial fluid leukocyte count, synovial fluid protein concentration, sedimentation rate, hematocrit, disease duration, or extraarticular manifestations. Levels of PAA in rheumatoid synovial fluid were significantly higher than in noninflammatory fluids (P < 0.03), and tended to be higher than in gouty fluids, al- -= though statistical significance was not achieved (P 0.08). High levels of PAA were detected in each juvenile rheumatoid synovial fluid, ranging from 1299 to 5100 units; these levels were significantly higher than those in noninflammatory (P< 0.005) or gouty (P < 0.02) fluids. 1 I 0 11,200- 9600 - '3 P 8 80000 6400- i? 4800- 0 0 8 3200 16001 0 RA JRA lnflommatory Figure 3. Units of platelet-activating activity (PAA) for 1.0 ml of partially purified synovial fluid from patients with adult rheumatoid arthritis (RA), juvenile rheumatoid arthritis (JRA), gout, and nonidammatory arthropathies. The broken line indicates 766 units, which is 2 standard deviations above the mean for noninflammatory fluids. SHAPLEIGH ET AL Table 2. Clinical characteristics and synovial fluid parameters for 13 adult rheumatoid patients Synovial fluid parameters Diagnosis Seropositive rheumatoid arthritis Seronegative rheumatoid arthritis Age Sex 81 46 45 74 50 F M M F 31 62 26 60 75 65 F M 59 61 Plateletactivating activity (units of PAA) c3 %/gm synovial fluid protein IgM RF (titer) < 50 Leukocyte count* 43 5 74 8 7 3 11,439 85 2,630 3,495 1:5,120 1:640 1 : 640 Negative Negative 16,615 12,470 9,600(65) F M F M 3,845 5,000 5,495 6,175 7,350 11,880 Negative Negative Negative 1: 160 1 : 2,560 Negative 13,680 8,840 9,770 5,000 10,233 8,614 SSO(35) 200(25) 4,000 (25) 40 40 18 4 16 3 18 26,200 (80) 16,600(97) 42 90 11 F < 50 M 4,760 Negative Negative 11,980 10,260 17,800(57) 1,250(48) 54 39 < 50 F 9,900(75) Disease duration ESRt (years) 13,150 (46) 10,800 5 1 1 6 Extraarticular manifestations None Nodules None Keratoconjunctivitis sicca None Nodules None None Nodule on chest x-ray None None None * Parentheses indicate percent polymorphonuclear leukocytes. t Erythrocyte sedimentation rate determined by the Westergren method. Synovial fluid samples were assayed for immune complexes by Clq binding and inhibition of ADCC (Table 3). Rheumatoid synovial fluids contained a significantly higher frequency of immune complexes than noninflammatory fluids (P < 0.03) as measured by Clq binding. For all synovial fluids studied, there was a correlation between levels of immune complexes as measured by the Clq binding qssay and units of PAA (rs = 0.40, sample size = 26, P < 0.05) (Figure 4).The level of PAA was elevated in each of 4 JRA synovial fluids, whereas Clq binding was increased in only one; conversely, units of PAA were increased in only one gouty fluid, whereas C l q binding activity was elevated in three of four. The ADCC inhibition assay was positive in all patient groups and did not correlate in frequency or titer with the C l q binding or platelet release assays. The possible modulation by IgM rheumatoid factor of platelet activation by IgG complexes in crude rheumatoid synovial fluid was investigated with purified SF-PAA and purified monoclonal IgM rheumatoid factor (mRF). SF-PAA was preincubated with buffer Table 3. Frequency of elevated levels of platelet-activating activity and immune complexes in synovial fluids Number of patients with elevated synovial fluid levels/total number studied Assay PAA units* C Iq binding assayt Inhibition of ADCC$ Adult rheumatoid arthritis Juvenile rheumatoid arthritis Gout Noninflammatory arthropathies 9/13 9/12 4/4 1/4 2/4 1/4 3/4 3/4 0/7 1/6 4/6 5/12 * Platelet-activating activity, elevated * t Elevated = > 23 pg/ml 0 ~ = > 766 units. aggregated IgG equivalent. Inhibition of antibody-dependent cellular cytotoxicity, elevated > 1.5 pg/ml aggregated IgG equivalent. = 2000 4000 .** I 6000 8000 UNKS OF PAA I - 10000 12000 Figure 4. Correlation between immune complex levels by C l q binding assay and units of platelet-activating activity (PAA) for synovial fluids from all patient groups (P < 0.05, Spearman’s rank correlation coefficient, r, = 0.40, sample size = 26). PLATELET-ACTIVATING ACTIVITY alone or with increasing concentrations of mRF at 37°C for 1 hour and then at 4°C for 1 hour. SF-PAA, in a final concentration of 0.021 mg/ml, stimulated a net I4C serotonin release of 37%. The percent inhibition of serotonin release with SF-PAA:mRF molar ratios of 128:1, 64:1, 12.8:1, 6.4:1, and 2:l were 0%, lo%, 45%, loo%, and loo%, respectively. In 3 consecutive experiments, 100% inhibition of serotonin release occurred at molar ratios ranging from 6.4:1 to 2: 1. Of interest was the finding that 3 of the 11 seropositive rheumatoid synovial fluids had low levels of PAA, and all three contained IgM rheumatoid factor, the titers ranging from 1:640 to 15,120. The remaining fluids all contained more than 766 units of PAA, and 6 of 8 had no IgM rheumatoid factor detectable in the synovial fluid (Table 2). DISCUSSION Previous studies have shown that the predominant platelet-activating activity in rheumatoid synovial fluid (SF-PAA) was an intermediate molecular weight complex of IgG. This purified material stimulated release of platelet I4C serotonin in a dose-response fashion, whereas crude synovial fluid did not (9). Thus a 5 ml concentrate of the 450,000 molecular weight material obtained by Sepharose 6B gel filtration of 1 ml of synovial fluid was used to assess PAA in synovial fluids of patients with inflammatory and noninflammatory arthropathies (Table 1). Additional minor peaks of PAA were recognized and found to contain less than 10% of the total PAA (Figure 1). The net percent release equal to that obtained with 0.08 units/ml thrombin was chosen to define one unit of PAA in order to standardize the assay from day to day and because I4C serotonin release with this dose was 44 & 11% (mean & SD) and fell on the dose-response curve for the synovial fluid PAA (Figure 2). Platelet-activating activity of greater than 766 units of PAA per ml was defined as elevated because this level was 2 standard deviations from the mean for the noninflammatory synovial fluids (Figure 3). Reproducibility of the assays was confirmed by serial assay of the 5 ml concentrate stored at 4°C which revealed that fluids with or without elevated levels of PAA retained their initial classification (Figure 2). Stability of PAA after freezing synovial fluid and storing it for up to 6 months at -70°C was demonstrated for one rheumatoid fluid which contained 11,880 units of PAA before and 10,000 units after freezing and for one gouty and one noninflammatory fluid which each remained negative. 805 Each of 7 noninflammatory and 3 of 4 gouty fluids contained <766 units of PAA, whereas 9 of 13 adult rheumatoid fluids contained elevated levels of PAA (Figure 3). This incidence is consistent with the Clq binding positivity of this group (Table 3 and Figure 4), and with reports of immune complexes in 7 1% to 95% of rheumatoid synovial fluids in which the Clq binding assay was used (22-24). Of the 4 rheumatoid patients with low levels of PAA, one was seronegative with recent onset of RA, and 3 were seropositive and had high titers of synovial fluid rheumatoid factor (Table 2), which would both inhibit SF-PAA and change its filtration characteristics. Salicylate therapy would not influence PAA because free or bound salicylate in synovial fluid would be excluded from the 5 ml concentrate by gel filtration. Synovial fluid PAA does not correlate with sedimentation rate, hematocrit, synovial fluid leukocyte count, synovial fluid protein, disease duration, or extraarticular manifestations in patients with adult rheumatoid arthritis. Although levels of immune complexes in serum, as measured by the Clq binding assay, correlate with the presence of extraarticular manifestations, synovial fluid levels do not (20). Each of the 4 JRA patients with seronegative oligoarticular disease had elevated levels of synovial fluid PAA, whereas C l q binding was negative in 3 of 4. Although l l of 5 l JRA patients had circulating immune complexes as measured by the C l q binding assay (25), immune complexes have not been reported in synovial fluid of patients with JRA. The presence of complexes in synovial fluid is suggested by the intraarticular acquired abnormalities of the classic complement pathway in approximately one-half of JRA patients, regardless of subtype or seropositivity (26). The levels of synovial fluid immune complexes assessed by Clq binding and units of PAA correlated (r, = 0.40, n = 26, P < 0.05) (Figure 4) for the whole group, possibly because the exceptions, namely JRA and gout, represented only 8 of the 26 patient fluids studied with both assays. Three of 4 gouty synovial fluids contained elevated levels of aggregated IgG equivalents by C l q binding, perhaps because urate crystals bind IgG (27) and activate complement (28); only 1 of 4 gouty synovial fluids contained elevated quantities of PAA. Activation of human platelets occurs via an Fc receptor as demonstrated by the stimulation of platelet aggregation and release of serotonin by aggregated IgG, IgG complexes, and aggregated Fc fragments but not aggregated F a b , fragments (6,7). Furthermore, platelet serotonin release by purified SF-PAA SHAPLEIGH ET AL 806 was inhibitable by monomeric IgG (9). The capacity of mRF to completely inhibit 14Cserotonin release by SFPAA in molar ratios of SF-PAA:mRF of 2:l to 6:l is compatible with the finding that IgG complex-induced platelet aggregation is inhibitable by cryoglobulin IgM components with rheumatoid factor activity (29). The findings of local production of IgM rheumatoid factor by rheumatoid synovium (30), the absence of free rheumatoid factor in rheumatoid synovial eluates with high levels of IgG complexes, even when these antibodies were present in serum (3 l), and the inhibition of SF-PAA-induced platelet serotonin release by purified mRF suggest that one role of IgM rheumatoid factor in the joint space is the blocking of IgG complexmediated inflammation. REFERENCES 1. Kunkel HG, 'Miiller-Eberhard HJ, Fudenberg HH, To- 2. 3. 4. 5. 6. 7. 8. 9. masi TB: Gamma globulin complexes in rheumatoid arthritis and certain other conditions. J Clin Invest 40: 1 17129, 1961 Schrohenloher RE: Characterization of the y-globulin complexes present in certain sera having high titers of anti-y-globulin activity. J Clin Invest 49501-5 12, 1966 Pope RM, Teller DC, Mannik M: The molecular basis of self-association of antibodies to IgG (rheumatoid factors) in rheumatoid arthritis. Proc Natl Acad Sci USA 7 1:517521, 1974 Hannestad K: Presence of aggregated yG-globulin in certain rheumatoid synovial effusions. Clin Exp Immunol 2:511-529, 1967 Winchester RJ, Agnello V, Kunkel HG: Gamma globulin complexes in synovial fluids of patients with rheumatoid arthritis: partial characterization and relationship to lowered complement levels. Clin Exp Immunol 6:689-706, 1970 Henson PM, Spiegelberg HL: Release of serotonin from human platelets induced by aggregated immunoglobulins of different classes and subclasses. J Clin Invest 52:12821288, 1973 Mueller-Eckhardt C, Liischer EF: Immune reactions of human blood platelets. I. A comparative study of the effects on platelets of heterologous antiplatelet antiserum, antigen-antibody complexes, aggregated gamma globulin and thrombin. Throm Diath Haemorrh 20:155-167, 1968 Myllyla G, Vaheri A, Penttinen K: Detection and characterization of immune complexes by the platelet aggregation test. 11. Circulating complexes. Clin Exp Immunol 8:399408, 197 1 Valone FH, Austen KF, Goetzl EJ: Identification of the platelet-activating activity in rheumatoid synovial fluid as an intermediate molecular weight complex of IgG. J Immunol 122:703-709, 1979 10. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ: Protein determination with the F o l k Phenol reagent. J Biol Chem 193:265-275, 1951 11. Virella G, Waller M, Fudenberg HH: Nephelometric method for determination of rheumatoid factor. J Immuno1 Method 22:247-251, 1978 12. Singer JM, Plotz CM: The latex fixation test. I. Application to the serologic diagnosis of rheumatoid arthritis. Am J Med 21:888-892, 1956 13. Zubler RH, Lange G, Lambert PH, Miescher PA: Detection of immune complexes in unheated sera by modified '251-CIq binding test: effect of heating on the binding of C l q by immune complexes and application of the test to systemic lupus erythematosus. J Immunol 1 16:232-235, 1976 14. Kammer GM, Schur PH: Binding of circulating immune complexes to human peripheral blood lymphocytes: effect of complement. Clin Immunol Immunopathol 10:202213, 1978 15. Kammer GM, Soter NA, Schur PH: Circulating immune complexes in patients with necrotizing vasculitis. Clin Immunol Immunopathol 15:658-672, 1980 16. Lewis EJ, Carpenter CB, Schur PH: Serum complement component levels in human glomerulonephritis. Ann Intern Med 75:555-560, 1971 17. Hunsicker LG, Ruddy S, Carpenter CB, Schur PH, Merill JP, Miiller-Eberhard HJ, Austen KF: Metabolism of third complement component (C3) in nephritis: involvement of the classic and alternate (properdin) pathway for complement activation. N Engl J Med 287:835-840, 1972 18. Gabriel A Jr, Agnello V: Detection of immune complexes: the use of radioimmunoassays with C l q and monoclonal rheumatoid factor. J Clin Invest 59:990-1001, 1977 19. Ropes MW, Bennet GA, Cobb S, Jacox R, Jessar RA: 1958 revision of diagnostic criteria for rheumatoid arthritis. Bull Rheum Dis 9:175-176, 1958 20. Brewer EJ, Bass J, Baum J, Cassidy JT, Fink C, Jacobs J, Hanson V, Levinson JE, Schaller J, Stillman JS: Current proposed revision of JRA criteria. JRA Criteria Subcommittee of the American Rheumatism Association, Section of the Arthritis Foundation. Arthritis Rheum (Suppl) 201195-199, 1977 21. Kellegren JH: The Epidemiology of Chronic Rheumatism. Vol. 11. Atlas of Standard Radiographs of Arthritis. Oxford, Blackwell Scientific Publications, 1963, pp 2-23 22. Zubler RH, Nydegger U, Perrin LH, Fehr K, McCormick J, Lambert PH, Miescher PA: Circulating and intraarticular immune complexes in patients with rheumatoid arthritis: correlation of lZ5I-CIq binding activity with clinical and biological features of the disease. J Clin Invest 57:1308-1319, 1976 23. Gupta RC, McDuffie FC, Huston KA, Tappeiner G, Meurer M, Jordan RE, Luthra HS, Hunder GG, Ilstrup D: Comparison of three immunoassays for immune complexes in rheumatoid arthritis. Arthritis Rheum 22:433439, 1979 PLATELET-ACTIVATING ACTIVITY 24. Halla JT, Volankis JE, Schrohenloher RE: Immune complexes in rheumatoid arthritis sera and synovial fluids. Arthritis Rheum 22:440-448, 1979 25. Rossen RD, Breuer EJ, Person DA, Templeton JW, Lidsky MD: Circulating immune complexes and antinuclear antibodies in juvenile rheumatoid arthritis. Arthritis Rheum 20:1485-1490, 1977 26. Rynes RI, Ruddy S, Spragg J, Stillman JS, Austen KF: Intraarticular activation of the complement system in patients with juvenile rheumatoid arthritis. Arthritis Rheum 19:161-167, 1976 27. Hasselbacher P, Schumacher HR: Immunoglobulin in tophi and on the surface of monosodium urate crystals. Arthritis Rheum 21:353-361, 1978 807 28. Naff GB, Byers PH: Complement as a mediator of inflammation in acute gouty arthritis. I. Studies on the reaction between human serum complement and sodium urate crystals. J Lab Clin Med 81:747-760, 1973 29. Wager 0, Penttinen K, Rasanen JA, Myllyla G: Inhibition of IgG complex-induced platelet aggregation by antiglobulin-active cryoglobulin IgM components. Clin Exp Immunol 15:393408, 1973 30. Smiley DJ, Sachs C, Ziff M: In vitro synthesis of immunoglobulin by rheumatoid synovial membrane. J Clin Invest 47:624, 1968 3 1. Munthe E: Relationship between IgG complexes and anti-IgG antibodies in rheumatoid arthritis. Acta Rheum Scand 16:240-256, 1970 Rheumatology and Metabolic Bone Disease The Harvard Medical School Department of Continuing Education will present a course entitled “Rheumatology and Metabolic Bone Disease” October 20-24, 1980 at the Parker House Hotel, Boston. Sponsored by The Massachusetts General Hospital, the course is designed for rheumatologists and internists with special interests in rheurnatology and metabolic bone diseases, with emphasis on recent advances in pathogenesis and treatment. Directors for the program are Drs. Dwight R. Robinson, Kurt J. Bloch, Stephen M. Krane, and John A. Mills. Tuition is $360 (U.S.), and attendance is limited. Harvard Medical School certifies that this continuing education program meets the criteria for approximately 3 1 credit hours in Category I of the Physician’s Recognition Award of the AMA. Application forms and additional information may be obtained from Harvard Medical School, Department of Continuing Education, 25 Shattuck Street, Boston, MA 02115.