A study of complement fixation by rheumatoid factor using a haemolytic assay system.

Complement fixation by rheumatoid factor (RF) in sera from rheumatoid arthritis patients has been investigated by means of a haemolytic assay system employing sheep erythrocytes (SRBC) coated with reduced and alkylated rabbit IgG anti-SRBC antibody. Haemolysis was almost invariably detected with RF-positive sera whereas haemolysis was not observed with RF-negative sera; It was evident from a marked increase in haemolysis, obtained following isolation of IgM-RF of high purity from several RF-positive sera, that the full in vitro complement-fixing ability and hence lytic potential of RF in serum is masked. Parallel observations were also made with synovial fluids. The inhibition of haemolysis by RF in sera and synovial fluids commonly involved a reduction in degree of lysis at high concentration of test material as well as a more generalized inhibition in overall percentage lysis. Complete replication of typical RF-positive serum and synovial fluid patterns of lysis on dilution was achieved by addition of heat-aggregated human IgG to isolated IgM-RF preparations and demonstrated that the two inhibitory effects are separable in terms of complement depletion and reduction of free rheumatoid factor antibody activity, respectively.     

Comparison of the behaviour of IgG and IgM anti-Forssman antibodies in agglutination, haemolysis and cytolysis

Three anti-Forssman sera, prepared after two, three and six courses of injections with horse kidney antigen, have been fractionated by DEAE-cellulose chromatography. The IgG and IgM antibody activities in the various fractions have been measured by agglutination of untreated and papain treated sheep red cells. They have also been tested by haemolysis of sheep red cells and by immune cytolysis of a continuous line of dog kidney cells.

The activity of IgG antibody measured by haemolysis and cytolysis are similar, but the IgM fraction shows much greater activity by haemolysis than by cytolysis.

It appears that IgM anti-Forssman antibody is inefficient in lysing nucleated cells with guinea-pig complement.

     

Differential effectiveness of complement fixed by IgG and IgM antibodies in the cytolysis of dog kidney tissue culture cells

Sensitizing and complement fixing properties of IgG (7S) and IgM (19S) rabbit anti-horse kidney (Forssman) antibodies have been studied. Agglutination, indirect antiglobulin and mixed antiglobulin techniques were used to determine the extent of antibody sensitization of sheep red cells and dog kidney tissue culture cells. Immune adherence and mixed conglutination were used to demonstrate the presence of sites of complement fixation on these sensitized cells. The relationship between the presence of fixed complement and the subsequent immune lysis of the test cells was examined.

Guinea-pig serum-complement appeared to be less well fixed by IgG than by IgM antibodies. However, the complement fixed by IgG antibody was shown to be highly lytic both for sheep red cells and dog kidney tissue culture cells. The fixation of complement by IgM antibodies resulted in weak and relatively ineffective cytolysis of dog kidney cells, but effective haemolysis of sheep red cells.

The sensitivity of the immune adherence test for detection of `natural' antibodies was also demonstrated.

     

The use of passive haemolysis in the quantitative estimation of anti-protein antibodies

The conditions for the establishment of a method for the quantitative assay of antiprotein antibodies by the use of passive haemolysis were studied by using an anti-BSA—BSA system as a model.

When haemolytic assaying of antibody was carried out under standard conditions with a fixed concentration of optimally sensitized red cells, in the presence of excess complement (10 C′H₅₀) reproducible titrations could be performed to detect 0.12 μg. N Ab with an error of ±10 per cent.

Factors affecting the specific lysis of optimally sensitized cells were studied and some parameters fixed. The amounts of coated erythrocytes and complement and the length of time of reaction were seen to influence the final degree of lysis given by a fixed amount of antibody.

The conditions of optimal sensitization of erythrocytes with protein antigen by the use of bis-diazotized benzidine (BDB) are indicated. It was found that, when conjugation was carried out at 0° with proper amounts of a standardized cell suspension, BDB and antigen, the resulting coated cells were highly sensitive to immune haemolysis and highly resistant to `spontaneous' lysis.

     

The demonstration of the involvement of antibody in the in vitro lysis of chicken erythrocytes by allergized spleen cells

The experiments described demonstrate that the in vitro lysis of chicken erythrocytes (CRC) by allergized mouse spleen cells (ASC) is an antibody-dependent phenomenon. Supernatants obtained from cultures of ASC were shown to contain antibody which could effect antibody-dependent cell-mediated cytotoxicity.

The cytolysis observed with ASC was inhibited by both an IgG fraction and a F(ab')₂ portion of rabbit anti-mouse IgG. No inhibition was obtained by pretreating the ASC with AKR anti-C3Hθ serum and guinea-pig complement.

     

The virulence for mice of strains of Escherichia coli related to the effects of K antigens on their resistance to phagocytosis and killing by complement

Strains of Escherichia coli with sufficient K antigen to resist killing by complement were poorly phagocytosed when injected intravenously into mice. Phagocytosis was markedly increased by anti-OK but not by anti-O sera. In contrast anti-K sera had little or no effect on the bactericidal reaction. This was not because K antigenic sites were scarce but may have been because their position was such that complement was activated at a distance from its substrate. Red cells coated with K antigen were poorly lysed by complement and anti-K serum, suggesting that the K antibody did not activate complement very effectively although again the sites may have been too superficial. The effect of K antigens on phagocytosis and complement killing or lysis could all be explained by their ability to impair protein binding.

Strains of E. coli rich in K antigen were resistant to phagocytosis and complement killing and were virulent for mice on intracerebral injection. The significance of K antigens in animal and human infections is discussed.

     

Quantitation of C3 subcomponents on red cells coated with complement in vitro

In order further to characterise and evaluate the reproducibility of human red cells coated with complement in vitro, the number of molecules of C3 subcomponents/red cell were determined by Scatchard analysis of equilibrium concentrations of bound and free antibody using ¹²⁵I-labelled goat anti-rabbit IgG. A 1:1 combining ratio was assumed. Red cells coated via the classical pathway had twice as much bound C3b and C3d as alternative pathway-coated cells. Assays using different anti-C3d sera gave different amounts of bound antigen, but results with any one antiserum versus one cell type were reproducible. Anti-C3d sera raised to C3d-tryp and to C3d-KAF detected significantly different amounts of bound C3d on the same cells. Both trypsinisation and serum KAF treatment of classical pathway-coated cells resulted in marked reduction of C3b molecules/cell (over 90% in both cases). Similar reduction in bound C3b was seen after trypsinisation of alternative pathway-coated cells, but serum KAF treatment of such cells had no significant effect. K₀ values were lower with anti-C3c than with anti-C3d. Anti-C3d K₀ values with the various cells coated with complement in vitro were not statistically different (approximately 10⁷ litres/mol), with the exception of trypsinised alternative pathway-coated cells (approximately 10⁸ litres/mol, the same order of magnitude observed with cells coated with C3d in vivo). A non-linear relationship between antiglobulin titre and antigen strength was observed. The minimal number of C3d molecules/red cell detectable by agglutination with the various anti-C3d sera ranged from 200 to 670 molecules. The minimal number of C3b molecules detectable by agglutination was approximately 9000 molecules/cell.     

The role of complement and gp120-specific antibodies in virus lysis and CD4+ T cell depletion in HIV-1-infected patients

The substantial virus lysis was induced by HIV-1-infected patient serum and normal human complement serum in the presence of purified patient IgG. Non-infected CD4+ T cells coated with the whole virus or with a recombinant HIV-1 envelope gp120 and sensitised with patient IgG were also shown to be susceptible to complement-dependent lysis. The serum level of complement regulatory protein in a fluid phase, the C1-esterase inhibitor, was significantly correlated with serum concentration of C1q-circulating immune complexes (P=0.0062), but inversely with CD4+ T cell count (P< 0.0001). Accordingly, the disease progression in HIV-1-infected patients was significantly correlated with the level of complement activation as determined by serum level of C1-esterase inhibitor (P=0.0001), and inversely correlated with CD4+ cell count (P< 0.0001) and gp120-specific antibody titre (P=0.0086). These results strongly suggest that the complement activation by gp120-specific antibodies play a very important role in virus clearance, but also in depletion of infected as well as gp120-coated non-infected CD4+ bystander T cells during the course of HIV-1 infection.     

Hepatitis C virus (HCV)-induced IgG-IgM rheumatoid factor (RF) complex may be the main causal factor for cold-dependent activation of complement in patients with rheumatic disease

A low serum complement level is commonly found in patients with rheumatic diseases. We evaluated 170 patients with such diseases to determine their serum levels of CH₅₀, C3 and C4 protein. Persistent hypocomplementaemia was found in 19 of those patients, particularly in those with systemic lupus erythematosus (SLE). Cold-dependent activation of complement (CDAC) was demonstrated in nine of the 19 (47.4%), and six of the nine patients demonstrated infection with HCV (66.7%). The nine patients that exhibited CDAC had nearly normal haemolytic complement activity when the sera were separated either at 37°C or in EDTA-treated plasma. Conversely, it markedly decreased, even to the point of being immeasurable, when the sera were separated at 4–21°C. No significant deficiency in C3 and C4 protein levels was found in these patients. Clinical parameters other than levels of anti-HCV antibody, transaminase, and RF were not influenced by CDAC. In an attempt to isolate the causal factor for CDAC, we isolated IgG fractions from the CDAC patients by using a protein G column, in which case precipitates were collected from the eluates. The precipitates were mixed with normal serum and incubated at 4–21°C for 18 h. A decrease in the level of CH₅₀ in normal serum was observed, which predominated (P<0.001) when precipitates from HCV-infected patients were used. This indicated CDAC was possibly interrelated to the precipitates of such patients. This precipitate was proved to contain IgM besides IgG. It is therefore possible that an HCV-related IgG complex or an IgG-IgM RF complex may be formed at low temperature and be involved in activating the complement system in vitro.     

Serum factors responsible for killing of Shigella

Eight strains of Shigella were tested for susceptibility to killing by seven normal human sera. Although there was a wide range of susceptibility between strains of bacteria, there was surprisingly little difference in the killing activity of individual sera and no relationship between antibody titres and killing capacity.

Bacteriolysis required small amounts of antibody, but as little as 0.02 mg of a 19S fraction from normal serum restored full killing capacity to 1 ml of antibody depleted serum. Neither 11S IgA nor Cohn fraction II restored the bacteriolytic ability. Both the early reacting complement sequence and the alternate C3 activating pathway appeared to participate in killing as indicated by the roles of C2 and C3PA. Killing occurred, but with reduced efficiency, when either of the two substances was missing. However, serum lacking both C2 and C3PA could no longer kill Shigella. Killing also required the presence of C3, and presumably some of the later components of complement are subsequently involved.

     

Solid-phase enzyme immunoassay or radioimmunoassay for the detection of immune complexes based on their recognition by conglutinin: conglutinin-binding test. A comparative study with 125I-labelled C1q binding and Raji-cell RIA tests

Bovine conglutinin was used in a solid-phase assay for the detection of immune complexes. In a first step, the tested serum sample is incubated in polypropylene tubes coated with conglutinin to allow C3-coated immune complexes to bind to solid-phase conglutinin. In a second step, the conglutinin-bound complexes are detected using an enzyme-conjugated or radiolabelled anti-immunoglobulin antibody.

The conglutinin-binding (KgB) test does not suffer from the interference of DNA, heparin or endotoxins. Its limit of sensitivity for aggregated IgG is 3 μg/ml undiluted human serum. Immune complexes prepared in vitro using tetanus toxoid, or DNA, and corresponding antibodies in human sera could be detected at various antigen/antibody ratios and at antibody concentrations lower than 8 μg/ml. The KgB test allowed for the detection of immune complexes in sera from patients with systemic lupus erythematosus, rheumatoid arthritis, idiopathic vasculitis, leprosy and leukemia. These sera were also tested using the ¹²⁵I-labelled Clq-binding activity (BA) test and the KgB test simultaneously, and a significant rank order correlation was observed. In patients with leukemia, a significant correlation was observed using three tests, KgB, ¹²⁵I-labelled Clq BA and Raji-cell radioimmunoassay (RIA).

Therefore, the KgB test appears as a simple and reproducible method, utilizing a very stable reagent, with a sensitivity and specificity comparable to the other tests studied and allowing for clinical application.

     

Studies of bactericidal activity to Escherichia coli of porcine serum and colostral immunoglobulins and the role of lysozyme with secretory IgA

Gel filtration and immune inhibition techniques were used to study bactericidal activities of IgG, IgM and IgA against smooth strains of Escherichia coli 0141 and 08 in sow serum and colostrum and post-colostral piglet serum. Bactericidal activity in sow sera was primarily associated with IgM and a low molecular weight IgG component, 7S IgG activity was less frequently observed. In colostral whey fractions and post-colostral piglet sera, in the absence of lysozyme, bactericidal antibody activity was associated with IgM and 7S IgG. In post-colostral serum bactericidal antibody was also attributable to a low molecular weight form of IgG.

IgA in serum from the sow and neonate showed no bactericidal activity, even in the presence of lysozyme, whereas in colostrum secretory 11S IgA had bactericidal activity, but only in the presence of complement and lysozyme.

     

Detection of antichlamydial immunoglobulin G and M antibodies by enzyme-linked immunosorbent assay

Chlamydia trachomatis causes a wide range of infections in adults and conjunctivitis and pneumonia in neonates. The complement fixation test for chlamydial antibody is broadly reactive, but possesses low sensitivity, whereas the microimmunofluorescence test is highly sensitive, but technically difficult to perform. A simple, rapid enzyme-linked immunosorbent assay (ELISA) has been developed for the measurement of immunoglobulin G (IgG) and IgM antibodies to C. trachomatis. Wells of microtiter plates were coated with Renografin-purified elementary bodies (serotype L2) grown in cycloheximide-treated McCoy cells, and serum antibody was detected with peroxidase-labeled goat antihuman IgG and IgM antibody. Of 41 sera tested from patients with lymphogranuloma venereum, pelvic inflammatory disease, cervicitis, or urethritis there was a 90 and 63% correlation of positive results for IgG and IgM, respectively, by microimmunofluorescence and ELISA. Of the positive correlates, ELISA titers were up to 128 times higher than microimmunofluorescence titers for IgG and IgM. The ELISA detected no false-positive results, but missed two positive results for IgG. Both of these sera were reactive against serotypes C and J, suggesting that the ELISA with LGV L2 antigen may not measure antibodies to serotypes within the C serogroup. The IgM ELISA detected 7 negative and 4 positive results not detected by the microimmunofluorescence test. Of four paired sera examined by ELISA, three showed a fourfold rise in IgG antibody titer, and one showed a twofold rise. Further evaluation of this ELISA will be required to determine how useful it will be in seroepidemiological studies and as a diagnostic tool.     

Correlation of disease activity and Clq-binding immune complexes with the neutrophil inclusions which form in the presence of SLE sera

Intracytoplasmic inclusions containing immunoglobulin (Ig) and complement (C3) are found in normal neutrophils (PMN) after incubation with sera from patients with SLE. These inclusions are believed to be immune complexes removed by phagocytosis from the SLE patients' sera in vitro. Similar inclusions were also noted in the circulating PMN from some patients with SLE. In the present study we have examined the relationship between the presence of intracytoplasmic inclusions and various clinical and laboratory features of SLE. Blood from forty-five patients with SLE was drawn and separated at 37°C. Fresh heparinized blood was also obtained from normal volunteers and allowed to stand for 90 min at 37°C. The buffy coat cells from both normal and SLE groups were removed, centrifuged, washed and examined (direct method) or incubated in the SLE sera for 90 min at 37°C (indirect method). Slides of washed cells were prepared in the cytocentrifuge, stained with fluorescein-conjugated goat anti-human IgG, IgM, IgA and C3 and examined under ultra-violet light.

By the direct method, 24% of patients had small intracytoplasmic inclusions in their neutrophils when stained for IgG suggesting that immune complexes were phagocytosed in vivo. None of twenty-one normal controls had similar inclusions. By the indirect method, 62% of SLE patients were positive for IgG, 15% for IgM, 8% for IgA and 31% for C3. None of the twelve normal controls were positive.

By the indirect method, PMN inclusions containing both IgG and IgM correlated with clinical activity (P<0·001), depressed serum complement (CH50, P=0·026; and C3, P<0·051), cryoglobulinaemia (P=0·014), anti-nDNA antibodies (P<0·001) and Clq-binding immune complexes (P=0·008). A suggestive correlation with granulocytopenia was also observed. The presence of inclusions containing IgG alone did not correlate with any of these parameters. C3 and IgM appeared to be mutually exclusive, i.e. neither was present simultaneously. These findings suggest (1) that normal PMN on exposure to SLE sera develop intracytoplasmic inclusions by phagocytosis of immune complexes, (2) the presence of such complexes correlates with a number of parameters of disease activity, particularly when IgG and IgM are both present and (3) such complexes may be phagocytosed in vivo as suggested by the presence of inclusions in vivo and contribute to a number of granulocyte disturbances seen in patients with SLE. These abnormalities in granulocyte function may be important, predisposing factors for infection in patients with active SLE.

     

Studies on passive haemolysis mediated by antiserum globulin antibodies

The concurrent action of complement and antiserum globulin (anti-SG) antibodies on red cells coated with an antigen and sensitized with the specific antibodies was studied. The occurrence of haemolysis in such systems is dependent on the haemolytic properties of the anti-SG antibodies used in the test. The haemolytic properties of the immunoglobulins that sensitize the coated cells have no apparent role in the development of the lysis.

The method of titration of antibodies by passive haemolysis can be rendered more sensitive and can be applied to detect antibodies that do not fix complement, if a suitable anti-SG serum is permitted to react with the immunoglobulins that have combined with the antigen coating the erythrocytes. Under standard conditions reproducible titrations can be performed to detect 0·009 μg N Ab/ml with an error equal to 10 per cent.

     

Properties of protective malarial antibody

Properties of protective malarial antibody have been studied in cultures of P. knowlesi giving average parasite multiplication rates of sixfold in 24 hours. Parasite growth was assessed by incorporation of [³H]-leucine into protein.

Immune serum has little effect upon the growth of intracellular parasites, but prevents reinvasion of red cells and inhibits the succeeding cycle of parasite development. Protective antibody is present in relatively low titre in immune sera even after long immunization and this may explain certain characteristic features of malarial immunity.

Protective antibody in the sera studied is associated with IgG and IgM; its action is not complement dependent, but requires at least two combining sites per molecule. Anti-malarial antibody has several features in common with viral neutralizing antibody.

     

The compatibility of Xenopus antibody and homoiothermic complement

Adult Xenopus laevis serum containing an anti-chicken erythrocyte antibody exhibits complement (C) activity by both the classical and the alternative pathways. Lytic activity is removed by heating serum to 50 degrees C for at least 20 min; it is also removed by polysaccharides and by 7.5 mM EDTA. The optimum temperature for Xenopus C activation is 25 degrees C. The ability of Xenopus antibody against chicken erythrocytes to co-operate with homoiothermic C in in vitro by lysis of chicken erythrocytes was tested. Removal of Xenopus C by heating resulted in apparent damage to the antibody, so C was removed by absorption with antibody/antigen complex. Xenopus antibody will apparently combine with guinea-pig and rat C to effect lysis of chicken erythrocytes; low levels of haemolysis were seen with human, rabbit and chicken sera, while horse and mouse sera were non-lytic. Xenopus C was able to restore the lytic activity of decomplemented rabbit anti-chicken erythrocyte serum.     

Complement-dependent cytotoxic antibodies to human T lymphotropic virus type I (HTLV-I)-infected cells in the sera of HTLV-I-infected individuals

To investigate whether HTLV-I induces the development of complement-dependent cytotoxic antibodies in humans, sera of asymptomatic HTLV-I carriers and of patients suffering from tropical spastic paraparesis/HTLV-I-associated myelopathy (TSP/HAM) or adult T cell leukaemia (ATL) were used in a cytotoxicity assay against a panel of target cells. This panel included uninfected cell lines (CEM, Jurkat, Molt and H9), cell lines chronically infected with HTLV-I (MT2, MT4, C91PL and HUT102), as well as lines H36 (H9 infected with HTLV-I), H9-IIIB (H9 infected with HIV_IIIB) and H9-MN (H9 infected with HIV_MN). HTLV-I⁺ sera induced lysis of H36 and of lines expressing HTLV-I antigens in the presence of rabbit complement, but did not lyse cells in presence of human complement. The HTLV-I⁺ sera also failed to lyse the HTLV-I⁻ lines and H9 cells, suggesting that lysis was specific for HTLV-I. H36 cell lysis was prevented by IgG depletion of the sera and by dialysis of rabbit complement against EGTA or EDTA. Rabbit complement-dependent cytotoxic antibodies were present in the sera of 14/14 HTLV-I-infected individuals; the highest titres were predominantly found in the sera of the TSP/HAM patients. Such antibodies were also detected in 5/5 individuals coinfected with HIV-1 and HTLV-I, although no cytotoxic antibody could be found against HIV-infected cells. Vice versa, sera of HIV-1-infected individuals did not exert a lytic effect in the presence of complement (of human or rabbit origin) against HIV-1- or HTLV-I-infected cells. Incubation of the sera of four HTLV-I-infected patients with HTLV-I env-specific synthetic peptides demonstrated that some of the complement-dependent cytotoxic antibodies recognized epitopes located on gp46 between amino acids 190 and 209. There is no correlation of rabbit complement-dependent cytotoxic HTLV-I antibodies with the development of disease.     

Complement mediated hepatocytes injury in a model of autoantibody induced hepatitis

Despite multiple reports on autoantibody-initiated complement activation in autoimmune hepatitis (AIH), how does the humoral immunity contribute to the pathogenesis of AIH remained unclear. In this report, by adoptively transferring a polyclonal rabbit anti-OVA antibody into Hep-OVA Tg mice in which OVA is selectively expressed on the surface of hepatocytes, we found that excessive complement activation initiated by the autoantibody overwhelmed the protection of intrinsic cell surface complement regulators, and induced hepatocytes injury both in vitro and in vivo. The anti-OVA antibody induced hepatic injury in Hep-OVA Tg but not WT C57BL/6 mice as assessed by serum ALT levels and liver histopathology. Immunohistochemical analyses showed that after the antibody administration, there was massive complement activation on anti-OVA IgG coated hepatocytes in Hep-OVA Tg mice, but not in WT mice. Consistent with these results, depleting complement by cobra venom factor (CVF) prior to antibody injections protected Hep-OVA Tg mice from anti-OVA IgG induced hepatic injury. In addition, treating Hep-OVA Tg mice with recombinant mouse decay accelerating factor, a native complement inhibitor, protected them from autoantibody induced hepatitis. These results suggest that complement could play a pivotal role in liver specific autoantibody mediated hepatocyte injury in AIH, and that complement inhibitors could be, in principle, developed as novel therapeutics against AIH.     

Characterization of two different agglutinators in the latex fixation test, occurring in normal human sera

Using a sensitive modification of the latex fixation test it is possible to detect a small agglutinating effect in about 60 per cent of normal human sera, after these have been heated for 30 minutes at 56°. This was shown to be caused by an IgM globulin with the properties of a rheumatoid factor. The factor is able to react with human IgG globulin and may represent an antibody to the IgG part of circulating antigen—antibody complexes. The heat treatment probably inactivates an inhibitor of the latex fixation reaction.

In addition all normal human sera give an agglutination reaction with IgG coated latex at incubation temperatures of 37° or lower. It was shown that these reactions are caused by a thermolabile, non-reducible component with a sedimentation constant of about 10. This component is probably identical with the complement component C'_1q. The agglutinating activity was found in the α₂—β₁ region after electrophoresis of untreated serum, but in the slow γ region after treatment of the serum with EDTA. This kind of agglutination may cause false positive reactions in latex tests which are carried out at 37° or less.