by rheumatoid sera and synovial fluid

Transcription

1 Annals of the Rheumatic Diseases, 1983, 42, Direct activation of neutrophil chemiluminescence by rheumatoid sera and synovial fluid R. GALE, J. V. BERTOUCH, J. BDLEY, AND P. J. ROBERTS-THOMSON From the Department of Clinical mmunology, Flinders Medical Centre, Bedford Park, South Australia 542 SUMMARY The majority of paired sera and synovial fluids from 21 patients ith rheumatoid arthritis produced a rapid chemiluminescent response hen incubated ith human neutrophils. Synovial fluid gave considerably higher responses than the paired serum specimen. n contrast little or no response as found ith paired sera and joint fluid taken from patients ith gout, psoriasis, and osteoarthritis and ith sera from healthy donors. A similar chemiluminescent response as observed hen neutrophils ere preincubated ith large aggregates of heated human gammaglobulin (HAGG), hich ere used as a model of immune complees. Smaller nonreactive aggregates of gammaglobulin became reactive after preincubation ith a purified monoclonal rheumatoid factor (mrf) hich had a high avidity for aggregated gg. The addition of this monoclonal rheumatoid factor also caused enhancement of chemiluminescence by rheumatoid sera. Further evidence suggesting that the active material found in these rheumatoid specimens contained compleed immunoglobulin as obtained by indirect immunofluorescence. Neutrophils developed intracellular immunoglobulin inclusions after preincubation in reactive rheumatoid sera but not ith nonreactive or normal sera. Hoever, activation of neutrophil chemiluminescence by rheumatoid specimens did not correlate significantly ith levels of rheumatoid factor or immune complees suggesting that the activating complees ere of a particular type. n conclusion e have shon the direct activation of neutrophil chemiluminescence by rheumatoid sera and synovial fluid and suggest that the activation is caused by large gg-containing immune complees. t is possible that this activation may have important implications in the immunopathogenesis of the rheumatoid inflammatory process. Rheumatoid arthritis () is an autoimmune disorder of uncertain immunopathogenesis, but immune complees have been strongly implicated in many of the clinical manifestations.' 2 t is possible that an important component of the inflammatory processes characteristic of this disease may be the result of direct neutrophil activation during the phagocytosis of these reactive immune complees both in the circulation and in the articular, pleural, and pericardial spaces.3`6 The resultant neutrophil activation may initiate the production and release of highly reactive superoide free radicals hich are likely candidates for causing subsequent tissue damage. There is some evidence to support this hypothesis in. Neutrophils taken from patients ith Felty's Accepted for publication 12 March Correspondence to Dr P. Roberts-Thomson. syndrome (FS) contain larger intracellular immunoglobulin inclusions and have reduced functional activity.3 Similar findings have been described folloing the incubation of normal neutrophils in sera from patients ith.4 Previous studies of the interaction beteen serum or synovial fluid from and FS patients have centred on the alterations to neutrophil functional capability.4 n this study e have used chemiluminometry to demonstrate the direct activation of normal neutrophils by rheumatoid sera and synovial fluid and provide evidence relating to the nature of this reactive material. Materials and methods Samples of serum and knee synovial fluid ere obtained from 31 patients ith joint diseases. There ere 21 patients ith definite or classical rheumatoid 158 Ann Rheum Dis: first published as /ard on 1 April Donloaded from on 24 November 218 by guest. Protected by

2 arthritis () defined according to the American Rheumatism Association criteria and 1 patients ith other joint diseases (non-). This latter group consisted of patients ith gout, psoriasis, and osteoarthritis. To patients ith Felty's syndrome ere defined by the presence of in association ith splenomegaly and neutropenia (leucocytes <2/mm3 (2 1O/l)). A single specimen of highly reactive pleural fluid as studied in more detail, as large volumes ere available. This specimen as obtained from an elderly male patient ith probable. His symptoms included an acute symmetrical polyarthritis ith pleural and pericardial effusions. Rheumatoid factor and antinuclear antibody ere both absent in the pleural fluid, but rheumatoid factor as present in the serum in lo levels. Neutrophils ere prepared from preservative-free heparinised blood obtained from group healthy individuals. The blood as diluted ith an equal volume of phosphate buffered saline (PBS) and the mononuclear cells removed by centrifuging through Ficol-Hypaque (Pharmacia). The pellet as resuspended in double the volume of 2 % detran in PBS and alloed to stand at room temperature for 45 minutes to sediment the red cells. The supernatant containing the neutrophils as removed, and contaminating red cells ere lysed by 5 minutes' incubation in Gey's solution. The preparation as adjusted to 5 16 neutrophils/ml in Eagle's MEM (Flo) containing 5 % fetal calf serum and maintained at 37C. Luminol (3-aminophthalhydraide, Koch Light Laboratories) stock solution as -56 M in dimethyl sulphoide. This as diluted 1:4 in MEM and prearmed to 37 C before use. Chemiluminescence (CL) as measured in an LKB luminometer (model 125) fitted ith a 37 C ater jacketed sample holder and lo-speed rotation of the reaction vessel to maintain continuous miing of the cell suspension. 2 Al of neutrophil preparation as incubated at 37 C for 5 min, ith 1,ul of the test serum, synovial fluid, or column fractions. 6 Al of luminol solution as then added and the rate of chemiluminescence activity recorded. The maimum chemiluminescent activity (CL ma) as defined as the highest recorded light output (in millivolts) integrated over one-second intervals. Heat-aggregated gg (HAGG) as prepared by heating Cohn fraction gammaglobulin (16 mg/ml (16 g/l), Commonealth Serum Laboratories), diluted 1:8 ith MEM at 63 C for varying periods of time up to 6 minutes. Filtration chromatography. Gel filtration as performed at room temperature ith either a cm Sephacryl S3 or a Sepharose 6B (Pharmacia) column ith PBS ph 7*3 ithout preservative as Direct activation of neutrophil chemiluminescence 159 eluant. An upard flo rate of 2 mvhour as maintained, and 6 ml fractions ere collected. ndirect immunofluorescent studies ere carried out on neutrophils preincubated for 1 hour in serum or synovial fluid at the same dilutions as used for chemiluminescence. The cells ere ashed and slides prepared ith a Shandon cytocentrifuge. The slides ere fied in 3 % formalin for 5 minutes, ashed tice in PBS, placed in acetone at -2 C for 5 minutes, and ashed tice ith ecess PBS. Fluorescein-labelled rabbit F(ab)' anti-immunoglobulin (Behring) diluted 1:2 in PBS as layered on the slide preparations and incubated at 4 C for 6 minutes. The slides ere then ashed, mounted, and eamined ith fluorescent microscopy under epiillumination. ncubation ith monoclonal rheumatoid factor (mrf). Serum samples or preparations of HAGG ere preincubated ith 1% v/v solution of a purified preparation of mrf (8.6 mg/ml (8-6 g/l)) for 6 minutes at 37 C before testing for neutrophil CL activity. The mrf as obtained from a patient ith a lymphoproliferative disorder, and its purification and properties have been previously described.8 Other immunological tests. mmune complees ere measured by the liquid phase Clq binding test, rheumatoid factor as measured by the sheep cell agglutination technique, and immunoglobulins ere measured by laser nephelometry as previously described.9 Results The maimum chemiluminescent activity (CL ma) of the paired serum and synovial fluid samples taken from 31 patients ith joint disease are shon in Fig. 1. Significantly raised levels of neutrophil chemiluminescence ere seen in the serum (p<5) and synovial fluid (p< 1) taken from patients ith rheumatoid arthritis compared ith samples taken fron non- patients. The highest CL ma responses ere produced by synovial fluid obtained from 2 patients ith Felty's syndrome and 5 patients ith severe active. Retrospective comparative studies of these results ith assays of Clq binding for immune complees and rheumatoid factor shoed no significant correlation. As large quantities of an inflammatory pleural fluid ere available from a patient ith probable, this specimen as studied in more detail. This fluid gave a high CL ma level (value = 315), contained large quantities of immune complees, and had greatly depressed levels of C3 and C4. The CL reactive material precipitated hen the specimen as incubated either at 4 C overnight or ith 3 % polyethylene glycol (PEG). This fluid as fractionated by Ann Rheum Dis: first published as /ard on 1 April Donloaded from on 24 November 218 by guest. Protected by

3 16 Gale, Bertouch, Bradley, Roberts- Thomson (U E uj uj -i Lu 66 4p 21 *@OO@@o os 8o Serum Joint Fluid Serum Joint Fluid Non Fig. 1 Neutrophil CL ma values (epressed in millivolts) for paired samples ofserum and synovial fluid taken from 21 patients ith classical or definite and 1 patients ith other joint diseases. The seven CL ma values >6 for knee synovial fluid varied over the range 7-6 and ere obtained from patients ith Felty's syndrome (recording levels of17 and 24 respectively) and 5 patients ith severe rheumatoid disease. Sepharose 6B filtration chromatography and the column fractions tested for CL activity and gg concentration. The peak CL activity as found in column fractions containing large molecular eight gg (MW >4 16 daltons), hile a small region of activity as seen in fractions eluting in the monomeric gg position (Fig. 2). Preparations of HAGG also induced neutrophil CL activity. Testing HAGG hich had been prepared by heating at 63 C for varying lengths of time indicated that a minimum heating time of 45-6 min as required before the preparations ere capable of inducing chemiluminescence. Aggregates induced by heating for shorter periods of time appeared to be poorly or nonreactive. After filtration chromatography of HAGG, fractions ere tested, and CL as again found to be restricted to those aggregates of large molecular sie (MW > 16 daltons) (Fig. 3). A monoclonal rheumatoid factor ith strong avidity for aggregated gg as preincubated ith sera from samples and controls. These preincubated samples shoed increases in CL activity greater than 1 times the preincubation levels (Fig. 4). Similar ELUTON VOLUME Fig. 2 Upper panel: The elution profile (% optical transmission) folloing sepharose 6B filtration chromatography of5 ml ofpleuralfluid. The elution position of the void volume (V1'), gm, and albumin are indicated. Middle panel: The gg concentration profile of the same column fractions shoing a large shelfof compleed gg etending into the void volume. Loer panel: The CL ma values obtained hen the column fractions ere incubated ith normal neutrophils. To peaks ofactivity coinciding ith the fractions containing large molecular eight compleed gg and monomeric lgg are demonstrated. Ann Rheum Dis: first published as /ard on 1 April Donloaded from on 24 November 218 by guest. Protected by

4 3 E 2 u uj C.) u. Vo gm t gg bp. i X P 6 a Direct activation of neutrophil chemiluminescence 161 E - F a.) U -Jo \C/ ~~~~b d ion ELUTON VOLUME Fig. 3 The elution profile ofa HAGG preparation after filtration chromatography on Sephacryl The CL ma readings obtained after incubating the column fractions ith normal neutrophils are indicated *-*. The elution position ofgm, monomeric gg, and the void volume (VO') are shon. CZ E CO D -J 7 SE Control 3 1 increases in CL activity ere observed hen preparations of HAGG hich had previously been nonreactive or poorly reactive ere preincubated ith mrf. The monoclonal reagent itself did not induce CL İndirect fluorescent eamination of neutrophils preincubated in CL reactive sera or synovial fluid shoed intracellular inclusions of immunoglobulin in more than 9% of the neutrophils. Less than 5 % of the neutrophils incubated in nonreactive control sera shoed positive staining. The antiserum used contained only F(ab). fragments, thereby avoiding the possibility of nonspecific staining of the neutrophil surface by binding to Fc receptors. Discussion The detection of a CL response hen normal neutrophils ere preincubated in samples taken from patients ith is direct evidence of neutrophil activation by those specimens. Particularly strong activation as observed ith specimens of synovial fluid obtained from patients ith severe or Felty's syndrome and ith a single inflammatory pleural fluid from a patient ith probable. These strong responses may reflect the site of highest concentration of the reactive material and hence the site of neutrophil activation in vivo. Detailed study of the single pleural fluid provided evidence relating to the nature of the reactive * S 21- -A- S 3 Control SE ith mrf Fig. 4 Neutrophil CL ma values observed folloing incubation of normal neutrophils in 13 and 9 control sera and the samples after preincubation ith mrf. Ann Rheum Dis: first published as /ard on 1 April Donloaded from on 24 November 218 by guest. Protected by

5 162 Gale, Bertouch, Bradley, Roberts- Thomson material. Cold and PEG induced precipitation and column filtration studies suggested that the reactive material contained compleed gg of large molecular eight. n addition a smaller CL peak as found eluting in the column fractions containing monomeric gg. The nature of the reactive material in this lo molecular eight region is unclear but might suggest a functionally active antineutrophil antibody. These observations may help to eplain the lack of correlation noted beteen the CL ma responses produced by the sera and synovial fluids and the amount of immune complees as measured by the Clq binding method and levels of rheumatoid factor. This lack of correlation suggests that the immune complees causing neutrophil activation may be different from those previously described in, hich are complement fiing and involve both gg and gm rheumatoid factor.1 The CL activity produced by pathological samples as mimicked by suspensions of HAGG hich ere used as a model of immune complees. Column chromatographic analysis of these preparations and the prolonged heating time required for the production of these reactive aggregates support the concept that large molecular sied gg aggregates are responsible for producing a CL reaction. Further indirect evidence that comple sie is important in determining CL reactivity as obtained using a mrf hich had a strong avidity for compleed gg. The addition of this reagent caused a dramatically amplified CL response. n addition the indirect immunofluorescent studies ith an antiserum directed against human immunoglobulin provide further evidence that the reactive material contains immunoglobulin. Fluorescent microscopy suggested that a correlation eists beteen the intensity of intracellular staining of the immunoglobulin inclusions in neutrophils and levels of CL reactivity produced by the same sera (unpublished). This area is being further investigated. The levels of CL induced by the more reactive rheumatoid specimens as demonstrated in this study are comparable ith the levels of CL produced during neutrophil phagocytosis of opsinised ymosan (unpublished). This suggests that the phenomenon is of some biological importance. Neutrophil CL is a measure of the generation of superoide and other 2 free radicals hich accompany phagocytosis.11 These high-energy metabolically reactive radicals have been strongly implicated as causative agents in the process of tissue damage.12 We propose that the direct neutrophil activation induced by immune comple containing rheumatoid sera and synovial and pleural fluid in vitro may also occur in vivo and as such have an important role in the immunopathogenesis of the inflammatory rheumatoid process. Note added in proof: Since submission of this paper e have noted similar findings by Starkebaum et al. 13 We thank Mrs M. Molnar for epert secretarial assistance and for typing the script. J. V. Bertouch is a greatful recipient of an NH and MRC postgraduate scholarship. References 1 Ziff M. Autoimmune processes in rheumatoid arthritis. Prog mmunol 1974; 5: Zubler R H, Nydegger U, Perrin L H, et al. Circulating and intra-articular immune complees in patients ith rheumatoid arthritis. J Clin nvest 1976; 57: Hurd E R, Andreis M, Ziff M. Phagocytosis of immune complees by polymorphonuclear leucocytes in patients ith Felty's syndrome. Clin Ep mmunol 1977; 28: Hoe G B, Fordham J N, Bron K A, Currey H L F. Polymorphonuclear cell function in rheumatoid arthritis and in Felty's syndrome. Ann Rheum Dis 1982; 4: Ugae K, Ziff M, Jasin H E. nteraction of polymorphonuclear leukocytes ith immune complees trapped in joint collagenous tissues. Arthritis Rheum 1979; 22: Hashimoto Y, Hurd E R. Human neutrophil aggregation and increased adherence to human endothelial cells induced by heat-aggregated gg and immune complees. Clin Ep mmunol 1981; 44: Doll N J, Wilson M R, Salvaggio J E. nhibition of polymorphonuclear leukocyte chemiluminescence for detection of immune complees in human sera. J Clin nvest 198; 66: Roberts-Thomson P J, Bradley J. A nephelometric study of the reaction of monoclonal rheumatoid factor ith heat aggregated gammaglobulin and sera from patients ith immune comple diseases. Clin Ep mmunol 1979; 37: Roberts-Thomson P J, Neoh S H, Bradley J, Milao S C. Circulating and intra-articular immune complees in rheumatoid arthritis: a comparative study of the Clq binding and monoclonal rheumatoid factor assays. Ann Rheum Dis 198; 39: Male D, Roitt M, Hay F C. Analysis of immune complees in synovial effusions of patients ith rheumatoid arthritis. Clin Ep mmunol 198; 39: Schadelin J, Schadelin R, Mandell G L. Chemiluminescence of phagocytic cells. CRC Crit Rev Clin Lab Sci 198; 13: Klebanoff S J. Oygen metabolism and the toic properties of phagocytes. Basic Revie. Ann ntern Med 198; 93: Starkebaum G, Stevens D L, Henry C, Gavin S E. Stimulation of human neutrophil chemiluminescence by soluble immune complees and antibodies to neutrophils. J Lab Clin Med 1981; 98: Ann Rheum Dis: first published as /ard on 1 April Donloaded from on 24 November 218 by guest. Protected by