Effect of IgM Fractions with or without Rheumatoid Factor Activity on Antigen-Antibody Reactions

The effect of isolated IgM fractions, with or without rheumatoid factor (RF) activity, on reactions between human albumin and rabbit anti-human albumin, human IgG and rabbit anti-human IgG, and tetanus toxoid and human anti-tetanus toxoid was assessed in the antigen excess zone. RF-active fractions increased and RF-negative IgM fractions decreased the amount of free antigen. The immune complexes that were affected by RF IgM fractions differed in composition from those affected by normal IgM fractions.     

Tetanus toxoid-anti-tetanus toxoid complexes: a potential model to study the complement transport system for immune complex in humans

Complement and its receptor on erythrocytes appears to play a physiological role in the elimination of large immune complexes (IC) in monkeys, and a similar system is likely to work in humans. Here we define a safe IC model which is suitable for clinical investigations. Soluble tetanus toxoid (TT)-human anti-TT (IgG) antibody complexes were prepared in large antibody excess. The size of the complexes was approximately 45 S. When incubated in normal human serum, 50% of the IC increased further in size, but remained soluble, and bound rapidly to human erythrocytes in vitro. This binding was shown to require intact classical pathway function. When injected into normal guinea-pigs a comparable proportion of IC bound immediately to blood cells (mainly to platelets). No platelet binding of IC occurred in C4-deficient guinea-pigs, but this binding was restored when C4 was supplied. Initial immune complex elimination was faster in C4 deficient than in C4-supplemented and normal guinea pigs. Thus classical pathway function appeared to be necessary for the normal processing, transport and elimination of TT-anti-TT complexes.     

C3b binding immune complexes and immunoconglutinins in human sera. Detection by enzyme-linked immunosorbent assay (ELISA)

An enzyme-linked immunosorbent assay (ELISA) designed to measure autoantibodies against C3b (immunoconglutinins: IK) also detects immune complexes (IC). Solid phase C3b, in addition to binding IK of IgG, IgM and IgA classes, bound aggregated human IgG, IgA and aggregated immunologically purified anti-tetanus toxoid antibodies as well as model complexes of tetanus toxoid-human anti-toxoid. Significant C3b binding IgG, IgM and IgA activities were seen in the sera of 20 SLE patients but not in sera from healthy blood donors. Ultracentrifugation analysis of two SLE sera revealed C3b binding IgG and IgA activities in both light (7S) and heavy (11S) fractions. This indicates simultaneous presence of IK and IC in these sera. On the basis of the known relationship between IK and IC formation we suggest that the solid phase C3b ELISA may be of value in evaluating immune reactions in patients.     

The binding of immune complexes to human red cells: complement requirements and fate of the RBC-bound IC after interaction with human phagocytic cells.

Soluble immune complexes (IC) are known to bind to human red blood cells (HRBC). Most authors have attributed this binding to the interaction between IC-bound C3b and a red cell CR1 receptor, but contradictory data has been published concerning the ability of IC to bind to HRBC in the absence of complement. Using soluble tetanus toxoid-rabbit anti-tetanus toxoid (TT-ATT) IC, we have shown that binding through the CR1 receptor takes place when IC are formed at antibody excess, while IC formed at antigen excess do not require complement for erythrocyte binding. Once absorbed to HRBC, IC are recognized by CR1 and/or Fc receptors on phagocytic cells. This interaction is not associated with red cell engulfment, but using radiolabelled S. aureus protein A as a probe, we have demonstrated the transfer of IC from HRBC to phagocytic cells. Such transfer without red blood cell (RBC) damage agrees with the postulated role of RBC in the elimination of soluble IC from circulation. However, we have also demonstrated that the interaction between HRBC-IC and phagocytic cells is associated with the release of mediators of inflammation. It is, therefore, not absolutely clear whether the interaction of RBC-adsorbed IC and phagocytic cells will always have beneficial consequences.     

Effect of C1q on the processing of immune complexes by human neutrophils.

We prepared immune complexes (IC) composed of human anti-tetanus toxoid IgG and tetanus toxoid, and examined the effect of C1q on the processing of IC by human neutrophils. Treating IC with increasing amounts of C1q enhanced the binding and phagocytosis of IC by neutrophils, unless the amounts of C1q added were less than those required to saturate the C1q binding sites of IC. With the increase of unbound excess C1q, the IC processing by neutrophils decreased. Superoxide anions generated during the processing of IC-C1q were entrapped in phagosomes and were not released from neutrophils. The C1q-dependent inhibition of IC processing by neutrophils was not observed when C1q-treated neutrophils were washed and allowed to react with IC, suggesting that the inhibition by excess C1q is due to the hindrance of IC-C1q binding to neutrophils by loosely bound C1q on neutrophils. The C1q-treated, washed neutrophils still showed enhanced responses to IC, suggesting that free C1q as well as the IC-C1q complex can prime neutrophils to enhance Fc receptor (FcR)-mediated cellular responses. Thus, C1q may have two effects on the processing of IC by neutrophils; firstly, it enhances FcR-mediated cellular responses, and secondly, it prevents superoxide anion-induced tissue damage by trapping superoxide anions in phagosomes.     

The complement subcomponent C1q mediates binding of immune complexes and aggregates to endothelial cells in vitro

The present studies were initiated to investigate whether soluble immune complexes, upon interaction with complement, can bind to endothelial cells. Human umbilical vein endothelial cells (HUVE) were incubated with purified human 125I-labeled C1q at 4 degrees C in RPMI-0.5% bovine serum albumin and assayed for binding. Optimal binding of 125I-labeled C1q to HUVE was reached within 2 h, and saturation of binding was found at concentrations of 5 micrograms/well input. The binding of 125I-labeled C1q was inhibitable with unlabeled C1q and by the collagenous region of pepsin-cleaved C1q. No inhibition was observed with the globular heads of C1q, suggesting that C1q binds to HUVE via the collagenous region of C1q. When HUVE were first reacted with various concentrations of C1q, washed and subsequently incubated with 125I-labeled aggregated human IgM (AIgM), binding of 125I-labeled AIgM to HUVE occurred depending on the dose of C1q. Only those aggregates of IgM which react with C1q in a solid-phase C1q binding assay were able to bind to HUVE presensitized with C1q. In addition it was shown that C1q mediated binding of aggregated IgG to HUVE. Furthermore, immune complexes (IC), that were prepared with bovine thyroglobulin (BTg) and rabbit anti-BTg, bound to C1q-preincubated HUVE. These studies suggest that localization of IC on endothelium can be enhanced following interaction of the IC with complement.     

Release of Immune Complexes Bound to CR1 on Erythrocytes; Suramin Inhibits Factor I in the Presence of EDTA

An experimental model was established in order to study the release of immune complexes (IC) bound by complement C3b receptors (CR1) on human erythrocytes (RBC). Soluble tetanus toxoid anti-tetanus toxoid complexes were incubated with RBC in the presence of autologous serum at optimal conditions for binding. The RBC carrying complement-opsonized complexes were incubated with appropriate serum reagents, and it was shown that factor I was required for release of the complexes, which occurred without loss of CR1. Suramin was, irrespective of factor I, found to induce release of CR1-bound IC in the absence of EDTA, whereas factor I-mediated release was inhibited by suramin in the presence of EDTA. EDTA probably interfered through a charge-dependent interaction. These observations are decisive for the interpretation of in vitro experiments involving these reagents. The combination of EDTA and suramin was found inappropriate for use in quantitative determination of in vivo CR1-bound IC.     

Quantitative studies of Fc receptors on human monocytes: characterization by binding of homologous and heterologous monomeric IgG and soluble immune complexes of different composition

The binding of 125I-labelled monomeric human and rabbit IgG (H-IgG, R-IgG) and rabbit IgG immune complexes (IC) to monocyte-enriched human peripheral blood cells had been investigated quantitatively. Scatchard plots at 4 degrees demonstrated that R-IgG bound to the same number of Fc receptors per cell (19,000) as H-IgG, but with a lower affinity (2.4 +/- 0.9 X 10(8)/l/mol and 3.5 +/- 1.1 X 10(8)l/mol, respectively). Inhibition studies demonstrated that the two ligands could mutually inhibit each other, H-IgG having higher inhibitory efficiency versus R-IgG than the reverse. It seems likely that R-IgG reacts with the Fc receptor for the homologous IgG, although with lower affinity. Binding of soluble R-IgG anti-bovine serum albumin (BSA) IC, prepared at molar antigen:antibody (Ag:Ab) ratio of 2:1 and 12:1, showed quite different behaviour, the IC binding with association constants almost 10-fold lower than the affinity of monomeric R-IgG, but binding six- to seven-fold as many IgG molecules per cell at saturation.     

Multispecificity of monoclonal rheumatoid factor

The study involved a monoclonal rheumatoid factor referred to as RF-AN, obtained by Steinitz and his associates from IgG-reactive human lymphocytes 'immortalized' by infection with Epstein-Barr virus. RF-AN combined with red blood cells (RBC) sensitized by either human or rabbit IgG antibodies. The monoclonal character of RF-AN strongly suggested that the same molecule of this factor combined with IgG of both species. Additional evidence for this contention was obtained from absorption and mixed agglutination experiments. RBC sensitized by either human or rabbit antibodies would remove serological activity of RF-AN for both human and rabbit IgG. RF-AN produced exclusively mixed agglutinates when reacted with a mixture of human RBC sensitized by human antibodies and chicken RBC sensitized by rabbit antibodies. Furthermore, it was shown that inhibition with aggregated human Fraction II gave strictly 'specific' results in that it abolished the reaction of RF-AN with RBC sensitized by human antibodies, but not by rabbit antibodies. This result was interpreted as indicating that RF-AN has a multispecific character, i.e., has separate combining sites for human and rabbit IgG. Interestingly, specific inhibition could not be achieved with aggregated rabbit Fraction II and this preparation affected reactivity of RF-AN with RBC sensitized by human antibodies as well as by rabbit antibodies.     

Neutralization of tetanus toxin by human and rabbit immunoglobulin classes and subunits

This investigation found that the human antibody class of importance in neutralizing tetanus toxin in mice was IgG, and that toxin neutralization was retained by the F(ab')2 and Fab' subunits of the human IgG class. Although human IgM and IgA classes appeared to neutralize tetanus toxin at very low levels, evidence was obtained that this neutralization was probably due to IgG contamination. Human Fabmu isolated from the IgM class did not neutralize tetanus toxin. Human antibodies of the IgG, IgM and IgA classes reacted with tetanus toxoid in the indirect haemagglutination (HA) test with IgG giving the highest HA titre. Rabbit antibodies of the IgG class also neutralized tetanus toxin, with neutralization being retained by the F(ab')2 and Fab' subunits of the rabbit IgG class. Absorption of several rabbit antisera to tetanus toxoid with goat-antirabbit Fc which is specific for absorption of IgG from antiserum, rendered them incapable of neutralizing tetanus toxin.     

Isolation of Immune Complexes by an Equine CIQ-Like Factor

A method for the isolation of circulating immune complexes by co-precipitation with an equine Clq-like factor is described. The Clq-like factor co-precipitated 80% of the ¹²⁵I-labeled aggregated human IgG (AHG) following dialysis against 0.05 M Tris-HCl, pH 8.1 compared to only 2% of the unaggregated human IgG (HG). When incubated with bovine serum albumin (BSA) anti-BSA soluble immune complexes (IC) prepared at antigen (Ag)-antibody (Ab) equivalence or in Ab excess, the Clq-like factor co-precipitated 100Z of the available complexes. Under conditions of 2x Ag excess, Clq-like factor was less efficient as demonstrated by co-precipitation of 40% of the available IC. Although the nature and specificity of binding of the Clq-like factor is unknown, its isolation and partial characterization are reported.     

The localization of immunoglobulin and immune complexes in lymphoid tissue

Aggregated and monomeric forms of human γ-globulin (HGG) were prepared by heating at 63°, ultracentrifugation and subsequent separation according to solubility in 0.62 M sodium sulphate. These two forms were injected intradermally into guinea-pigs'' ears and their distribution in the draining auricular nodes determined at different times following injection by staining cryostat sections with fluorescein labelled anti-HGG. Monomeric HGG showed no precise localization; aggregated HGG localized rapidly in the phagocytic macrophages of the sinuses and medulla and after a few hours'' delay in the germinal centres in a dendritic pattern, the latter persisting for up to 4 weeks. With doses of less than 10 μg, aggregated HGG was not seen in the medulla but germinal centre staining was readily visible, possibly due to a concentrating effect. Prior injection of a large dose of monomer either locally or systemically did not alter the pattern of staining produced by subsequent injection of aggregated HGG.Aggregated human serum albumin, colloidal carbon and streptococcal cell walls did not localize in germinal centres in the same way.Monomeric rabbit IgG anti-HSA injected alone did not localize, but when combined with HSA in antigen excess to form soluble immune complexes it localized in germinal centres.It is concluded that germinal centres contain receptors, probably at cell surfaces, for IgG aggregated by mild heat or by complexing with antigen but not for unaltered native IgG, and it is suggested that this may be a means of disposal of aggregated or complexed IgG.     

Neutrophil Fcγ and complement receptors involved in binding soluble IgG immune complexes and in specific granule release induced by soluble IgG immune complexes

We examined the effect of soluble IgG immune complex (IC) characteristics on the binding of IC to human neutrophils and IC-induced specific granule release of neutrophils via Fcγ receptors (CD16 and CD32) and complement receptors (CR1 and CR3). A set of soluble IgG IC varying in size, IgG subclass, antigen epitope density and complement (C) incorporation were formed between 5-iodo-4-hydroxy-3-nitrophenacetyl (NIP) coupled to bovine serum albumin (BSA) and chimeric mouse-human anti-NIP monoclonal antibodies (mAb) of all four IgG subclasses. High and low epitope density IC of all four IgG subclasses induced specific granule release with C, but in the absence of C only IgG1 and IgG3 IC were functionally active. The Fcγ and C receptors responsible for IgG IC-induced specific granule release and IC binding were determined using mAb specific for the ligand binding sites of CD16, CD32 and CR3, and recombinant soluble CR1. Each defined IC displayed a unique pattern of receptor preference, dependent upon subclass and antigenic epitope density. IC binding and IC-induced specific granule release was not mediated by the same receptor, or combination of receptors. High and low epitope density IgG3 IC binding and induction of specific granule release was mediated predominantly via CD16. Other IC subclasses bound differently, i.e. IgG1 IC used CD16 and CR3; IgG2 and IgG4 predominantly used complement receptors; but all three induced specific granule release via CD32. In vivo these results may translate into differential activation of neutrophils by soluble IC dependent upon their characteristics, leading to subtle nuances in the etiology, pathology and control of the immune response in IC-related diseases.     

Application of human erythrocytes to a radioimmune assay of immune complexes in serum.

An immune adherence receptor exists on the surface of primate erythrocytes, and has been characterized as a receptor for the activated third component of complement (C3b). We have applied human red blood cells (RBCs, blood group O) to a sensitive determination of complement-fixing, soluble immune complexes in serum. The method involved the binding of immune complexes with RBCs in the presence of complement and the detection of cell-bound IgG molecules by radiolabelled anti-human IgG antibodies. Since the binding of RBCs with monomeric IgG was minimal, cell-bound IgG molecules were taken as representing immune complexes. When aggregated human gammaglobulin (AHG) was used as a model of immune complexes, as little as 5 μg dissolved in 1 ml of normal human serum were detected. The binding of RBCs with AHG was inhibited in EDTA solution where the classical complement pathway could not be activated. The RBC radioimmune assay was successfully applied to the determination of soluble immune complexes in pathological serum samples obtained from the patients with systemic lupus erythematosus and those with fulminant Type B hepatitis. False-positive results by autoantibodies against RBCs could be excluded by performing a Coombs test and by comparing the binding in the presence of complement with that in EDTA solution. The ubiquitous availability of RBCs coupled with a high sensitivity would allow the RBC radioimmune assay to be added to the battery of previous methods to determine immune complexes in the serum.     

Immunological studies of human placentae: subclass and fragment specificity of binding of aggregated IgG by placental endothelial cells.

Fluorescein-conjugated heat-aggregated human IgG binds to endothelial cells of foetal stem vessels in cryostat sections of normal, full-term human placentae. No binding was observed using native human IgG of heat-aggregated human albumin, IgM or IgA2. No inhibition of binding of heat-aggregated human IgG was observed by pre-treatment of placental tissue sections with native IgG or non-aggregated Fc fragments. The binding was blocked using heat-aggregated Fc fragments prepared from IgG1, IgG2, IgG3 and IgG4 myeloma proteins, but not with heat-aggregated human light chains, Fab and F(ab)2 fragments of human IgG, or with heat-aggregated human IgM and IgA2. It is suggested that the placental endothelial cell receptor for aggregated IgG may function to keep immune complexes from entering the foetal circulation.     

Soluble immune complexes binding to human monocytes and polymorphonuclear leucocytes.

Soluble human serum albumin anti-albumin immune complexes (IC) have been shown to bind to freshly isolated human peripheral blood monocytes and polymorphonuclear leucocytes (PMN) in vitro. Binding sites on both cell types were saturable and specifically inhibited by heat aggregated IgG1 and IgG3 subclasses. PMN contained a greater number of binding sites than monocytes although the affinity was similar for both cell types. The enhanced binding of IC by both cell types observed after incubation at low pH (pH 6.0) was a consequence of increased affinity of the PMN binding site and an increase in the number of sites in monocytes. Binding of IC by both cell types was inhibited by normal human serum. Enhanced IC binding associated with enhanced affinity and number of sites was observed in PMN and monocytes preincubated in suspension with trypsin. However, monocytes exposed to trypsin while adherent to microexudate coated flasks demonstrated a marked increase in affinity without any change in the number of sites.     

Study of anti-DNA antibodies prepared by DNA cellulose or Cibacron blue chromatography

We prepared anti-DNA antibodies from sera of lupus patients by either DNA cellulose or by Cibacron blue chromatography. Eluates from both columns were studied with respect to recovery of IgG, recovery, purification and specificity of anti-DNA activity. An attempt was made to raise rabbit anti-idiotypic antibodies against both eluates. DNA cellulose chromatography--if DNA leakage was prevented--yielded 58% recovery and 58-fold purification of the anti-DNA activity present in the original purified IgG sample. 1% of loaded IgG was recovered. Cibacron blue chromatography yielded 32% recovery and 1.1-fold purification of the anti-DNA activity. 29% of loaded IgG was recovered. Eluates of Cibacron blue were not pure as shown by their high binding activity against an unrelated antigen, tetanus toxoid. Eluates from DNA cellulose were pure and did not show anti-tetanus toxoid activity. Rabbit anti-idiotypic antibodies could be raised only against eluates of DNA cellulose suggesting that the eluates of Cibacron blue did not contain enough idiotypes to induce anti-idiotypic antibodies. The characterization of the rabbit anti-idiotypic antibodies showed that it contained two populations, one against site-specific idiotypes and the other against framework idiotypes. Anti-DNA antibodies prepared by Cibacron blue had idiotypes similar to those prepared by DNA cellulose. The present study demonstrates that DNA cellulose chromatography--if leakage of DNA is prevented--can yield excellent recovery and purification of anti-DNA activity. Anti-DNA antibodies prepared by DNA cellulose were enriched and could induce anti-idiotypic antibodies in rabbits. Tube chromatography on Cibacron blue yielded poor recovery and minimal enrichment of anti-DNA activity.     

Characterization of the soluble immune complexes that are detected by three different techniques

Three different tests which are based on different principles were used for the detection of soluble immune complexes (IC): (i) a PEG precipitation test, which is based on the solubility characteristics of IC; (ii) a solid-phase C1q binding assay, which is based on the complement binding property of IC, and uses peroxidase-linked anti-human IgG to detect the bound IC (C1q-ELISA); and (iii) the indirect granulocyte phagocytosis test (IGPT) which is based on the Fc R- and possibly the C3 R-binding of IC. Using heat-aggregated IgG (A-IgG) as a model for soluble IC all three tests showed a linear relation with the amount of A-IgG. The C1q-ELISA and the IGPT had a detection limit of less than 1 microgram/ml while the PEG test only detected quantities of more than 10 micrograms/ml. However, when using artificially produced soluble IC, which were prepared from human antibodies (ab) of different specificities and their respective antigens (ag) i.e., (i) tetanus toxoid, (ii) Helix pomatia hemocyanin (HPH), and (iii) dsDNA, and which consisted of the two components in a wide range of ag/ab ratios, distinct results were obtained with the three tests. Thus demonstrating that results obtained with A-IgG as a model for soluble IC can not simply be extrapolated to the behavior of real complexes in IC detection assays. No matter which ag was used, the composition of the IC, i.e., the ratio in which ag and ab were present, appeared to be the crucial factor for detectability in the different tests. The C1q-ELISA can detect IC over a wide range of ag/ab ratios, while it is particularly sensitive for IC formed in slight ag excess. The IGPT in contrast primarily detects, and is highly sensitive for, IC formed in ab excess. The PEG test appears to detect IC with freshly bound complement only. Another interesting finding has to be mentioned here: when increasing amounts of dsDNA were added to a SLE serum containing anti-DNA ab, the IC that had been detectable in the native serum with the IGPT completely disappeared, thus demonstrating that these complexes did consist of DNA and anti-DNA.     

Control of the immune complex-complement interaction by protein H of the alternative complement pathway and the natural inhibitor heparin

The potential of the negative regulatory protein H of the alternative complement pathway convertase and of heparin in modulating the complement-dependent capacity of fresh serum to inhibit immune complex precipitation (CIICP) between bovine serum albumin (BSA) and rabbit anti-BSA as well as tetanus toxoid (TT) and human anti-TT was assessed. Additions of purified H to serum to increase the intrinsic concentration of this protein by 80% (BSA-anti-BSA system) and 190% (TT-anti-TT system) resulted in an inhibition of CIICP by 50% and 60%, respectively, whereas further increase of the amount of H lead to a decrease of its inhibitory activity. A similar effect was observed with heparin: at a concentration of 400 U/ml a 90% inhibition of CIICP in the TT-anti-TT system was obtained which diminished at higher heparin concentrations. The effect of H on C3 deposition to immune aggregates was assessed through its influence on C3b-mediated immune adherence hemag-glutination; factor H dose-dependently suppressed such hemagglutination induced by aggregated human IgG or preformed TT-anti-TT complexes when added to the immune complex-fresh serum mixture at the outset but not after 45 min of the 37 °C incubation period which means that H inhibited more likely decoration of immune complexes with C3b than it did inhibit the interaction of C3b-coated immune complexes with erythrocytes. This suppressive effect of H was reversed by the simultaneous addition of the activating protein B. Complement-mediated binding of tritiated C3 to latex-bound human IgG was assessed and H was found to dose-dependently inhibit such binding with a maximal inhibition of 37% at a H concentration of 7 μg/ml.     

Proatherogenic and proinflammatory properties of immune complexes prepared with purified human oxLDL antibodies and human oxLDL

Immune complexes (IC) prepared with human low density lipoprotein (LDL) and rabbit LDL antibodies induce foam cell transformation of human macrophages and activate the release of proinflammatory mediators by human macrophages and THP-1 cells. Because the affinity of human oxidized LDL (oxLDL) antibodies is lower than that of rabbit antibodies, IC formed with human antibodies could have limited pathogenic potential. Immune complexes prepared with human oxidized LDL (oxLDL) and purified human oxLDL antibodies (predominantly of the IgG1 and IgG3 isotypes) were presented to THP-1 cells using two protocols previously described in studies of the properties of LDL-IC prepared with rabbit antibodies. OxLDL/human oxLDL antibody IC immobilized by adsorption to red blood cells (RBC) induced the release of significantly higher levels of TNF from THP-1 cells (872-313 pg/ml) than oxLDL adsorbed to RBC (461-75.6 pg/ml) and caused a higher degree of cholesterol ester accumulation in the same cells (5.4-0.77 in cells incubated with IC-coated RBC vs 1.99-1.16 in oxLDL-coated RBC). Insoluble IC prepared with oxLDL/human oxLDL antibody were even more effective in promoting intracellular accumulation of cholesterol in THP-1 cells (total cholesterol = 53.8-13.5 and cholesterol esters = 24.0-7.2 mg/l in THP-1 cells incubated with insoluble IC (200 micrograms) vs total cholesterol = 32.4-8.2 and cholesterol esters = 7.7 +/- 2.8 micrograms/l in THP-1 cells incubated with an identical concentration of oxLDL) and also induced the release of TNF. Thus we have demonstrated that IC prepared with human oxLDL and human oxLDL antibodies have the same atherogenic and proinflammatory properties as IC prepared with human LDL and rabbit LDL antibodies. This strongly supports the concept that modified LDL-IC present in circulation and/or tissues play an important pathogenic role in arteriosclerosis.