Complement C4-derived monocyte-directed chemotaxis-inhibitory factor. A molecular mechanism to cause polymorphonuclear leukocyte-predominant infiltration in rheumatoid arthritis synovial cavities

To reveal the mechanism of the lesser infiltration of monocytes in synovial cavities with rheumatoid arthritis despite the presence of chronic inflammation, the synovial fluid from 15 rheumatoid arthritis patients was analyzed with respect to leukocyte chemotaxis. The synovial fluid possessed strong chemotactic activity to polymorphonuclear leukocytes but rather suppressed one to monocytes. The synovial fluid contained two different inhibitory activities in monocyte chemotaxis. One, which also suppressed polymorphonuclear leukocyte chemotaxis, was identified as alpha 1 protease inhibitor. The other, with molecular weight of 8 kd, possessed the specificity to monocytes and shared the antigenicity with complement C4 but not with C3 or C5. A similar inhibitor was generated in normal human plasma when the classical pathway of the complement system was initiated with aggregated human IgG, while it was not when alternative pathway was initiated with zymosan. The small size factor in the synovial fluid, apparently derived from C4, seemed to be a cyto-directed factor that might block an early part of signal transduction system of monocytes in the chemotaxis. After removal of the small-size inhibitor, the synovial fluid exhibited chemotactic ability to monocytes. Therefore the apparent C4-derived factor might play a key role in the polymorphonuclear leukocyte-predominant infiltration in the synovial fluid of rheumatoid arthritis.     

Interleukin 2 inhibitor in synovial fluid.

Since evidence for the presence of IL-2 activity in rheumatoid synovial fluid is conflicting, we have assayed IL-2 activity in synovial fluid from patients with rheumatoid arthritis (RA) and other articular diseases (OAD). Using the IL-2-dependent murine T cell line CTLL, IL-2 activity was not demonstrable in synovial fluid tested at concentrations ranging from 50% to 0.02%. There was an inhibitory effect on IL-2 activity in the bioassay of synovial fluid from 16 of the 22 patients with RA and 15 of the 16 with OAD. This inhibitory activity was heat-labile, precipitable by ammonium sulphate, reversible with excess IL-2 and was not significantly altered by preincubation of synovial fluid with CTLL. The mean inhibitory activity of synovial fluid from patients with RA was significantly reduced in comparison with that of synovial fluid from patients with OAD. Sera also had an inhibitory effect on IL-2 activity; however sera from patients with RA were less inhibitory than control sera but were more inhibitory than sera from patients with systemic lupus erythematosus. The deficiency in synovial fluid of an inhibitor of IL-2 activity may be relevant to the pathogenesis of RA.     

C5a-inhibitor deficiency in peritoneal fluids from patients with familial Mediterranean fever

Normal peritoneal fluid contains an inhibitor of neutrophil chemotaxis that acts by antagonizing the complement-derived chemotactic anaphyllatoxin C5a. The inhibitor resembles a substance previously described in synovial fluids and is a protein with a molecular weight of approximately 40,000 as determined by gel filtration. In contrast, levels of inhibitory activity in peritoneal fluids from five patients with familial Mediterranean fever were decreased to less than 10 per cent of those found in normal subjects. Gel filtration of peritoneal and synovial fluids from these patients did not yield any fraction with inhibitory activity. We suggest that C5a-inhibitor deficiency in joint and peritoneal fluids from patients with familial Mediterranean fever may have a role in the pathogenesis of the inflammatory attacks characteristic of this disease.     

Synovial fluid inhibits killing of Staphylococcus aureus by neutrophils.

Serum in the extracellular environment promotes neutrophil bactericidal activity apart from its opsonizing properties. We examined the effect of non-inflammatory osteoarthritic synovial fluid on serum-mediated neutrophil killing of Staphylococcus aureus. This was done to evaluate the effect of synovial fluid on neutrophil bactericidal activity independent of opsonin concentration. With an initial inoculum of 5 X 10(6) CFU/ml, 1.47 +/- 0.14% bacteria survived after 120 min of incubation with 10% serum and neutrophils. In contrast, 4.07 +/- 0.33% bacteria survived after incubation in serum plus synovial fluid (P less than 0.001). This inhibitory effect was directly related to the concentration of synovial fluid in the incubation mixture. Increasing the concentration of synovial fluid resulted in an increased percent survival. Studies utilizing preopsonized bacteria and radiolabeled organisms demonstrated that synovial fluid did not interfere with opsonization or phagocytosis. Intracellular bactericidal activity was assayed separately from phagocytosis by utilizing a brief ingestion period followed by the removal of extracellular bacteria by either differential centrifugation or lysostaphin treatment. The reincubation of cells and associated bacteria with serum or serum plus synovial fluid revealed that synovial fluid significantly inhibited the promoting effect of serum on neutrophil bactericidal activity. After 60 min of incubation with 10% serum, 13.0 +/- 1.2% bacteria survived, whereas 21.5 +/- 2.3% survived after incubation in serum plus synovial fluid (P less than 0.005). Superoxide production was not affected by the presence of synovial fluid. These findings suggest that the inhibitory effect of synovial fluid is due to an interaction between synovial fluid and the serum factors that promote intracellular killing.     

The chemoattractant activity of rheumatoid synovial fluid for human lymphocytes is due to multiple cytokines

The majority of synovial fluids from 29 rheumatoid arthritis patients were strongly attractive for normal blood lymphocytes judged by assays of polarization and collagen gel invasion. While rheumatoid synovial fluids contained IL-15, IL-8, monocyte chemotactic protein-1 (MCP-1) and macrophage inflammatory protein-1α (MIP-1α) at levels sufficient to attract lymphocytes, inhibition of the activity of any single cytokine using specific antibody did not abolish the activity of the fluid. However combinations of anti-cytokine antibodies used together (anti-IL-15+anti-MCP-1; anti-IL-8+anti-MCP-1 or +anti-MIP-1α) inhibited most of the activity, suggesting that attraction of lymphocytes by the fluids is due to a combination of attractants. Blood lymphocytes required activation by overnight culture to respond optimally, while rheumatoid synovial tissue lymphocytes responded to synovial fluids without a requirement for a period of culture. Lymphocytes derived from rheumatoid synovial fluids were poorly responsive to locomotor stimulants. Most of the responding cells from blood mononuclear cell fractions were T lymphocytes of the CD45RO isotype. Incubation in the presence of cyclosporin A or corticosteroids inhibited the response of lymphocytes to the fluids, but the presence of non-steroidal anti-inflammatory drugs (NSAIDs) and other agents used in therapy of the patients from whom the fluids were taken had no inhibitory effect.     

Selective inhibition of lymphocyte responsiveness to phytohaemagglutinin in patients with Reiter's syndrome.

The influence of synovial fluid from four patients with Reiter's syndrome on lymphocyte responsiveness was studied. On synovial fluid was found which specifically depressed the responsiveness of lymphocytes from patients with Reiter's syndrome to the non-specific mitogen phytohaemagglutinin (PHA). The ratio of the responsiveness of lymphocytes cultured in the presence of foetal calf serum (FCS), compared to those incubated in the Reiter's synovial fluid, was used as a measure of the depression induced by the Reiter's synovial fluid. The mean ratio for eight normals stimulated with PHA was 0-70 (range 0-35-0-96), while for eight patients with Reiter's syndrome, it was 0-13 (range 0-07-0-19). Similar studies done with concanavalin A (con A) showed no difference between lymphocytes from normals (0-73) or patients with Reiter's syndrome (0-67). Chromatography of the Reiter's synovial fluid on a Sepharose 4-B column resulted in the separation of three major fractions, one of which exhibited the inhibitory activity. When this active fraction was absorbed with Reiter's lymphocytes, a loss of the inhibitory activity of the fraction was seen. A similar absorption with normal lymphocytes had no effect. These studies demonstrate that a factor present in the synovial fluid of a patient with Reiter's syndrome reacted specifically with lymphocytes from patients with Reiter's disease and not with lymphocytes from normals. The interaction of this factor with lymphocytes from patients with Reiter's syndrome inhibited the responsiveness of these lymphocytes to PHA but not to con A.     

Degradation of Thymidine to Thymine by Rheumatoid Arthritis Synovial Tissue Eluate

A study was made of the effect of rheumatoid synovial tissue eluate on phytohaemagglutinin (PHA) stimulation of peripheral blood lymphocytes. It was found that rheumatoid synovial tissue eluate caused marked inhibition of 3H-thymidine incorporation by lymphocytes stimulated with PHA. Synovial tissue from traumatic joints had no effect. The inhibitor was a heat-sensitive and nondialysable substance. The inhibitory effect of PHA stimulation was diminished by thymine. In thin-layer chromatography it showed the capacity to convert thymidine to thymine. Because thymine is not incorporated by DNA-synthesizing cells, this report explains why negative results are obtained when lymphocytes are stimulated by rheumatoid synovial tissue using 3H-thymidine incorporation as an indicator of lymphocyte activation.     

Detection of interleukin 8 biological activity in synovial fluids from patients with rheumatoid arthritis and production of interleukin 8 mRNA by isolated synovial cells

The presence of neutrophils in the synovial joint of patients with rheumatoid arthritis (RA) is thought to be due to the activity of chemotactic factors released by activated cells in the joint. We have shown in this report, for the first time, the abundance of one such factor, interleukin 8 (IL 8), in the synovial fluid of patients both with RA and other non-RA joint diseases, and the spontaneous production of IL 8 mRNA by RA synovial cells in culture. There was no correlation between the levels of chemotactic activity and IL 8 protein, suggesting that other factors with similar neutrophil chemotactic activity are also present in the synovial fluid exudate. In support of this concept neither the level of chemotactic activity nor IL 8 protein levels correlated with neutrophil or leukocyte infiltration, indicating that the mechanism of migration into the inflammatory environment of the joint is complex. Such migration is likely to be due to a number of chemotactic signals in addition to IL 8, which may either synergize with, or inhibit, the action of IL 8.     

Detection of a complement-derived chemotactic factor for tumor cells in human inflammatory and neoplastic effusions.

A chemotactic factor for neoplastic cells can be generated in vitro by incubating human C5 or C5a with leukocytic or pancreatic lysosomal enzymes and is also detectable in experimental inflammatory exudates. The authors therefore sought evidence for the existence of this factor in human effusions. Using the Boyden chamber assay, they detected chemotactic activity for MB-MDA-231 human breast carcinoma cells and Walker ascites tumor cells in human inflammatory and neoplastic exudates, including ascites, pleural effusions, synovial fluids and cerebrospinal fluids. Chemotactic activity was not found in transudates, normal cerebrospinal fluid, or normal serum. Human ovarian adenocarcinoma cells from one of the effusions migrated toward autologous ascites and towards the C5-derived chemotactic factor that had been prepared in vitro. In gel filtration the chemotactic factor behaved generally as a molecule having a molecular weight of approximately 6000 daltons. The activity was blocked after incubation with antiserums directed against C5 but not by antiserums directed against C3 or C4. In vitro, chemotactic activity for tumor cells could be generated by incubating extracts of exudate cells with autologous plasma or with purified C5. The authors conclude that a chemotactic factor for tumor cells can be formed in human effusions and that this factor has properties similar to those of a previously described C5-derived chemotactic factor.     

Urinary trypsin inhibitor (UTI) and fragments derived from UTI by limited proteolysis efficiently inhibit tumor cell invasion

We investigated the effects of purified human urinary trypsin inhibitor (UTI) and fragments derived from UTI by proteolysis on the invasive potential of ovarian cancer cells (HOC-I) and gestational choriocarcinoma cells (SMT-ccl) using anin vitro reconstituted basement membrane invasion assay. These cells express cell-associated plasmin and functional uPA receptors that are partially occupied by ligands. SMT-ccl cells, which express threefold higher levels of cell-associated plasmin activity than HOC-I cells, showed approximately twofold increase in their invasive potential. For the invasion assay, HOC-I cells were primed with exogenous plasminogen, but SMT-ccl cells were not. Human leukocyte elastase (HLE)-digested UTI (22 kDa fragment; UTI-22) inhibited plasmin practically with the same strength as native UTI. Trypsindigested UTI (20 kDa fragment; UTI-20), however, did not inhibit plasmin significantly. Treatment of cells with UTI or UTI-22 reduced the incidence of tumor cell invasive capacity, whereas the inhibitory effect of UTI-20 was not remarkable. The inhibitory effect on tumor cell invasion was dose-dependent and non-toxic; moreover, it was not mediated by inhibition of the tumor cell chemotactic response or of cell attachment to matrigel. These results indicate that inhibition of the proteolytic enzyme plasmin specifically reduced the invasive capacity of tumor cellsin vitro.     

Hydroxyapatite associated arthritis in a patient undergoing chronic haemodialysis

A patient with uraemia undergoing chronic haemodialysis developed a unilateral effusion of the elbow joint. The chalky-white synovial fluid obtained from the joint contained many crystals that were identified as hydroxyapatite by infrared spectroscopy. The joint effusion showed chemotactic activity for healthy human neutrophils. Chemotactic activity of the lipopolysaccharide(LPS)-activated patient's serum was low compared to that of pooled healthy human sera, but the patient's serum did not have the inhibitor for the chemotactic factor. These findings suggest that the production of chemotactic factor in this patient's serum was decreased. The possible combined effects of the uraemic state and crystal-induced inflammation on the chemotaxis of neutrophil leukocytes are discussed.     

CHEMOTACTIC ACTIVITY FOR POLYMORPHONUCLEAR AND MONONUCLEAR LEUKOCYTES IN RHEUMATOID SYNOVIAL FLUIDS

In order to why polymorphonuclear leukocytes (PMNs) are predominant and mononuclear leukocytes (MNLs) are few in rheumatoid synovial fluids, chemotactic factor(s) for PMNs and MNLs were studied in the synovial fluids of rheumatoid arthritis (RA-SF) and osteoarthritis (OA-SF) using both Boyden's and agarose methods. The RA-SF showed strong chemotactic activity for human peripheral blood PMNs compared with non-rheumatoid OA-SF. The chemotactic activity for PMNs was well correlated with the number of PMNs in RA-SF, suggesting that it was a natural mediator for PMN emigration into rheumatoid joint cavity. The major chemotactic factor for PMN in RA-SF was of apparent molecular weight of 14,000 and its activity was suppressed to less than 10 percent by anti-C5a antibody, but it failed to show any anaphylatoxin activity which was an attribute of C5a. It was, therefore, suggested to be C5a-like molecule but not C5a itself. The possibility that the factor may be a C5a des-Arg was discussed. On the contrary, the chemotactic activity for MNLs was not found neither in RA-SF nor OA-SF. These findings may explain the fact that PMNs are predominant in rheumatoid synovial fluids.     

Chemotactic activity for polymorphonuclear and mononuclear leukocytes in rheumatoid synovial fluids

In order to why polymorphonuclear leukocytes (PMNs) are predominant and mononuclear leukocytes (MNLs) are few in rheumatoid synovial fluids, chemotactic factor(s) for PMNs and MNLs were studied in the synovial fluids of rheumatoid arthritis (RA-SF) and osteoarthritis (OA-SF) using both Boyden's and agarose methods. The RA-SF showed strong chemotactic activity for human peripheral blood PMNs compared with non-rheumatoid OA-SF. The chemotactic activity for PMNs was well correlated with the number of PMNs in RA-SF, suggesting that it was a natural mediator for PMN emigration into rheumatoid joint cavity. The major chemotactic factor for PMN in RA-SF was of apparent molecular weight of 14,000 and its activity was suppressed to less than 10 percent by anti-C5a antibody, but it failed to show any anaphylatoxin activity which was an attribute of C5a. It was, therefore, suggested to be C5a-like molecule but not C5a itself. The possibility that the factor may be a C5a des-Arg was discussed. On the contrary, the chemotactic activity for MNLs was not found neither in RA-SF nor OA-SF. These findings may explain the fact that PMNs are predominant in rheumatoid synovial fluids.     

Human alveolar macrophages release a factor that inhibits phagocyte function

Human alveolar macrophages release in vitro a factor that inhibits both random migration and chemotaxis of human polymorphonuclear neutrophils (PMN). This factor is not cytotoxic and is recovered in culture supernatants of alveolar cells from most nonsmoking normal subjects. The inhibitor can be detected 30 min after cell cultures are established and is still produced after 24 h in culture. Its release was inhibited by cycloheximide. When supernatants are separated by molecular sieving (I-60 Waters HPLC column), most of the inhibitory activity is recovered in the low-molecular-weight fractions of the chromatogram (less than 1,000 D). The inhibitor has a broad spectrum of activity against known chemoattractants in that it reduces significantly the chemotaxis of PMN induced by the formyl peptide FMLP, by the complement fragment C5a, and by leukotriene B4; it also decreases the chemotactic activity associated with a monocyte-derived interleukin 1 preparation and the chemotactic activity derived from alveolar macrophage culture supernatants. The inhibitory factor is partially heat labile, is sensitive to aminopeptidase M, and is nonpolar. Both phorbol myristate acetate (PMA) and FMLP-induced superoxide release by PMN are diminished significantly in the presence of this inhibitory factor (p less than 0.01 for PMA and p less than 0.05 for FMLP). The inhibitor also reduces monocyte chemotaxis but has no effect on monocyte random migration. Finally, studies with [3H]FMLP indicate that this inhibitor does not act at the site of receptor binding on PMN. Thus, human alveolar macrophages can release in vitro both neutrophil chemotactic factors and an apparent neutrophil-inhibiting factor that may modulate positively and negatively the movement and the respiratory burst of neutrophils in the alveolar space.     

Human neutrophil chemotactic response to group A streptococci: bacteria-mediated interference with complement-derived chemotactic factors.

The influence of M protein on the capacity of group A streptococci to generate neutrophil chemotactic activity in normal human serum was examined. Incubation of serum with M- bacteria for up to 10 min led to the production of chemotactic activity. In contrast, incubation of serum with M+ bacteria did not elicit serum chemotactic activity over a 1-h period, even though complement was activated to completion. Further experiments revealed that both M+ and M- bacteria could inhibit the chemotactic activity of serum preexposed to zymosan. However, the M+ bacteria possessed a 130-fold-greater inhibitory capacity in this regard than the M- bacteria. This antichemotactic property was not detectable in the fluid phase of serum incubated with bacteria, thereby ruling out neutrophil-directed effects. Treatment of the bacteria with trypsin resulted in the release of the inhibitory molecule, suggesting that proteins are involved in its maintenance at the cell surface. However, the resistance of the chemotactic factor inactivator to pepsin and trypsin indicated that the protease-sensitive M protein was not involved. These results demonstrate a heretofore uncharacterized activity of group A streptococci that may contribute to virulence through modulation of the host chemotactic response.     

Effect of ibuprofen and diclofenac on the chemotaxis induced by substance P and transforming growth factor-β on human monocytes and polymorphonuclear cells

The neuropeptide substance P and the cytokine transforming growth factor-β are potent chemotactic factors for monocytes or polymorphonuclear cells. They are present in synovial fluid of arthritic patients, and participate in the pathogenesis of arthritis. We investigated, in vitro, the effect of two non-steroidal anti-inflammatory drugs, ibuprofen and diclofenac, on the chemotactic effect of substance P and transforming growth factor-β at concentrations that can be present in the synovial fluid of arthritic patients. Both drugs decrease the chemotaxis induced by substance P and transforming growth factor-β, at concentrations that can be easily reached in the synovial fluid during therapy. This event could be involved in the effect of some non-steroidal anti-inflammatory drugs on the development and progress of arthritic disease.     

Studies on the first component of complement (C1) and the inhibitor of C1 esterase in rheumatoid synovial fluids

The specific activity of C1 (haemolytic units per unit of C1q or C1s protein) was two to three times lower in synovial fluid of patients with sero-positive rheumatoid arthritis (RA) than in non RA or sero-negative RA fluids. These fluids contain a high ratio of C1s inhibitor to C1s. Experiments with radiolabelled C1s established that the C1s inhibitor in the synovial fluids of these patients is functional as it binds radiolabelled C1s. At the high ratios observed, it is possible that a complex of C1s and C1s inhibitor forms in these synovial fluids and alters the activation of the complement system. The potential for excess C1s inhibitor to slow or diminish complement utilization was shown by adding purified C1s inhibitor to serum in order to simulate the conditions in joint fluid (i.e. a high ratio of C1s inhibitor to C1s). Subsequent whole complement, C1 and C4 activites after incubation with selected immune complexes were higher as compared to controls without added inhibitor.     

Migration of blood and synovial fluid neutrophils obtained from patients with rheumatoid arthritis.

The unstimulated random migration and the serum-induced chemokinesis of neutrophils obtained from the peripheral blood of patients with rheumatoid arthritis (n = 19) was not different from those of controls (n = 20). However, neutrophils obtained from the joint fluid of rheumatoid patients (n = 10) demonstrated a reduced serum-induced chemokinesis which was correlated with the amount of immune complexes present in the synovial fluid. The chemotactic response of peripheral blood neutrophils from subjects with rheumatoid arthritis taking aspirin (n =11) was increased while that of those rheumatoid subjects not taking aspirin (n = 8) was the same as controls. It is concluded that although there is no impairment of the in vitro migratory capacities of peripheral blood neutrophils obtained from patients with rheumatoid arthritis, neutrophils obtained from synovial fluids exhibit a marked defect in chemokinesis which may be related to the ingestion of immune complexes within the joint space.     

Modulation of neutrophil functions by activated platelet release factors

Platelets activated with physiological agonists, such as thrombin, ADP, or collagen, released products able to modulate neutrophil functions. In particular, platelet supernatant contained an inhibitor of superoxide anion generation induced by phorbol ester and a chemotactic factor for human neutrophils. The proteolytic digestion of platelet supernatant completely abrogated chemotactic activity without interfering with the inhibitory effect, indicating the presence of different molecules involved in the modulation of different neutrophil functions. This was further confirmed by the pretreatment of platelets with aromatic benzamidine which abolished chemotactic activity, but did not affect superoxide inhibition of neutrophils. This report provides evidence for interaction of platelets and inflammatory cells, suggesting that platelets are able to induce accumulation of neutrophils and control their respiratory burst, which also has a critical role in tissue damaging in inflammation.     

Inflammatory cytokine production induced by an analogue of muramyl dipeptide MDP-Lys(L18) in rat macrophage cultures and dog synovial fluid

To examine the involvement of cytokines in the mechanisms of N²-[(N-acetylmuramoyl)-L-alanyl-D-isoglutaminyl]-N⁶-stearoyl-L-lysine, MDP-Lys(L18)-induced arthritis, we analyzed interleukin-1 (IL-1), tumor necrosis factor (TNF), colony-stimulating factor (CSF), and neutrophil chemotactic factor (NCF) by bioassays in the rat macrophage-conditioned medium (Mø-CM) stimulated by MDP-Lys(L18) in vitro and the synovial fluid from dogs treated subcutaneously with MDP-Lys(L18) for 14 days in vivo. The dog showed arthritis characterized by swelling of the knee joint, increased synovial fluid and thickened synovial membrane, and a single subcutaneous injection of MDP-Lys(L18) was previously shown to induced synovitis in rat tarsal joint. IL-1, TNF, CSF, and NCF activities in Mø-CM were increased by MDP-Lys(L18), while only NCF activity was detected in the dog synovial fluid. Partial purification procedures revealed that NCF in Mø-CM was not leukotriene B₄ but a protein having heparin-affinity, and, in addition, immuno-reactive IL-8 was evident to be in Mø-CM. The NCF activity in the dog synovial fluid was not inhibited by dialysis, showing that NCF is a protein substance, possibly a chemokine. These results suggest that MDP-Lys(L18) produces a chemokine, such as IL-8, which recruits neutrophils to the synovial membrane for subsequent development of synovitis in rats and dogs.