Binding of immunoglobulin G aggregates and immune complexes in human sera to Staphylococci containing protein A.

Using the Cowan I strain of Staphylococcus aureus, we compared the binding properties of human monomeric immunoglobulin (Ig)G and oligomeric or complexed IgG. Heat-aggregated IgG served as a model for complexed IgG and heat-killed, formalin-fixed S. aureus (StaphA) as a cellular receptor for IgG, in determining the parameters for oligomeric and monomeric binding. Because of its capacity for multipoint attachment, complexed IgG binding was favored over monomeric IgG binding, and this preferential binding was demonstrated kinetically in equivalent forward rates of binding but in a much slower rate of release from StaphA receptors. From binding studies, we determined which conditions maximize complexes IgG binding and minimized monomeric IgG binding and applied them to the development of an assay for aggregated IgG and immune complexes in human sera. The StaphA binding assay that was devised is quantitative, sensitive, and not complement dependent. It is relatively unaffected by factors such as heparin, complement fixation, native antibodies, and immunoglobulin concentrations, but is affected by the presence of rheumatoid factors. It compares favorably with two other complement-dependent assays of immune complexes, the 125I-Clq binding assay and the Raji cell assay, in terms of sensitivity and the size of immune complexes detected. Studies on the potential of the assay for detecting, isolating, and characterizing immune complexes in biological fluids are presented.

Novel Fc gamma receptor I family gene products in human mononuclear cells.

Unlike the human Fc gamma RII and Fc gamma RIII families, which exhibit considerable diversity at both the nucleic acid and protein levels, the human Fc gamma RI family has only a single recognized product expressed as a 70-kD cell surface receptor with high affinity for monomeric IgG (hFc gamma RIa1). Using both polymerase chain reaction-based amplification and Northern hybridization, we document multiple interferon-gamma-inducible hFc gamma RI RNA transcripts in human mononuclear cells and neutrophils. The sequences of two of these Fc gamma RI related transcripts indicate that they are alternatively spliced products of a second Fc gamma RI family gene, termed Fc gamma RIB. The cDNA derived from the larger of these transcripts, termed hFc gamma RIb1, encodes a surface molecule that is not recognized by Fc gamma RI specific monoclonal antibodies when transfected into COS-7 cells. Unlike the interferon-gamma-inducible hFc gamma RIA gene product, hFc gamma RIb1 does not bind monomeric IgG with high affinity. However, both hFc gamma RIa1 and hFc gamma RIb1 do bind aggregated human IgG. Previously unrecognized diversity within the hFc gamma RI family includes an interferon-gamma-inducible, putative low affinity Fc gamma receptor that may play an important role in the mechanism by which Fc gamma receptors participate in the humoral immune response.

Identification of bacterial antigens in circulating immune complexes of infective endocarditis.

The presence of circulating immune complexes (IC) in patients with infective endocarditis has been well documented but the contributions of host and bacterial components to these IC have not been defined. To study this question, IC were isolated from serum of a patient with Streptococcus faecalis endocarditis by differential polyethylene glycol precipitation and competitive binding to staphylococcal protein A. A rabbit antiserum raised against the purified IC had reactivity by crossed immunoelectrophoresis primarily with an antigen derived from the cytoplasm of the infective organism. The antigen was a protein with a 12,000-dalton molecular mass. In situ radiolabeling of the IC bound to the protein A demonstrated a component of the same molecular mass as the bacterial antigen recognized by the antiserum. The patient serum had multiple antibody specificities reactive with bacterial antigens, including the antigen recognized by the rabbit anti-IC antiserum. These techniques for isolation and characterization of circulating IC may have value in the study of IC diseases in which the inciting antigens are not known.

A novel polymorphism of FcgammaRIIIa (CD16) alters receptor function and predisposes to autoimmune disease.

A novel polymorphism in the extracellular domain 2 (EC2) of FcgammaRIIIA affects ligand binding by natural killer (NK) cells and monocytes from genotyped homozygous normal donors independently of receptor expression. The nonconservative T to G substitution at nucleotide 559 predicts a change of phenylalanine (F) to valine (V) at amino acid position 176. Compared with F/F homozygotes, FcgammaRIIIa expressed on NK cells and monocytes in V/V homozygotes bound more IgG1 and IgG3 despite identical levels of receptor expression. In response to a standard aggregated human IgG stimulus, FcgammaRIIIa engagement on NK cells from V/V (high-binding) homozygotes led to a larger rise in [Ca2+]i, a greater level of NK cell activation, and a more rapid induction of activation-induced cell death (by apoptosis). Investigation of an independently phenotyped normal cohort revealed that all donors with a low binding phenotype are F/F homozygotes, while all phenotypic high binding donors have at least one V allele. Initial analysis of 200 patients with SLE indicates a strong association of the low binding phenotype with disease, especially in patients with nephritis who have an underrepresentation of the homozygous high binding phenotype. Thus, the FcgammaRIIIa polymorphism at residue 176 appears to impact directly on human biology, an effect which may extend beyond autoimmune disease characterized by immune complexes to host defense mechanisms.

Clinical and microbial features of prosthetic joint infection.

A one-year experience with prosthetic joint infection, in which 63 cases were identified, is reviewed. Thirty cases (48 percent) were early infections, in the first postoperative year, and 33 cases (52 percent) were late, occurring more than one year after implantation. Pain was the predominant symptom, but clinical clues suggesting infection were frequently absent, with fever in 43 percent and leukocytosis in only 10 percent. The radiographic appearance was more frequently abnormal in late infections (67 versus 37 percent, p less than 0.02). Staphylococci were predominant organisms, constituting 59 percent of prosthetic joint infections, and S. epidermidis was the predominant species in both early and later infections. Of the hematogenous infections, 11 of 13 occurred in the group with late infections; these were mostly nonstaphylococcal . Antigenic proteins of S. epidermidis were characterized by gel electrophoresis, but no infection-specific antigens could be identified when patient serum was compared with normal samples. Precipitating antibodies to the extracellular proteins of S. epidermidis were present in 50 percent of patients with S. epidermidis prosthetic joint infections, 27 percent of patients with nonstaphylococcal infections, 20 percent of patients with S. aureus infections, and 11 percent of normal subjects. In view of the increasing importance of prosthetic joint infection, further study of the pathogenesis of the infection and the host immune response is warranted.

Hypocomplementemia with low C1s-C1 inhibitor complex in systemic lupus erythematosus.

Ninety-three serum and plasma samples from 45 patients with systemic lupus erythematosus were analyzed for the complex formed by C1s and its inhibitor, as well as for C3, C4, C4a desarginine, and staphylococcal protein A-bound immune complexes. There were statistically significant correlations between C1s-C1 inhibitor complex and CH50, between C1s-C1 inhibitor complex and C4, and between C1s-C1 inhibitor complex and C4a desarginine. Serial studies were performed on 24 patients over a period of 6 months. Seven of 21 patients with hypocomplementemia had persistently normal levels of C1s-C1 inhibitor complex, 7 had transiently abnormal levels of C1s-C1 inhibitor complex, and 7 had sustained abnormal levels of C1s-C1 inhibitor complex. Two of 3 pregnant patients with normal levels of complement had abnormal levels of C1s-C1 inhibitor complex. Staphylococcal protein A-bound immune complexes demonstrated no correlation with any of the complement assays. Complement activation, as measured by C1s-C1 inhibitor complex, is often a transient phenomenon in systemic lupus erythematosus patients with persistent hypocomplementemia.

Human Fc gamma RIII (CD16). Isoforms with distinct allelic expression, extracellular domains, and membrane linkages on polymorphonuclear and natural killer cells.

Human Fc gamma RIII (CD16), a low-affinity receptor expressed on several different cell types, has a polymorphism on polymorphonuclear cells (Fc gamma RIIIPMN) identified by the NA1 and NA2 markers. Inasmuch as this polymorphism has functional consequences, an understanding of the structural biology of Fc gamma RIII may provide important insight into receptor function. We analyzed Fc gamma RIIIPMN by SDS-PAGE and found that receptor from individuals allotyped for either NA1 or NA2 contained only one protein after removal of N-linked glycosylations (19 and 21 kDa respectively) whereas receptor from NA1/2 individuals contained both bands. Because some reports indicate that digestion of Fc gamma RIII on NK cells (Fc gamma RIIINK) with N-glycanase also results in two bands on SDS-PAGE, we investigated Fc gamma RIIINK to explore the possibility of a corresponding allelic polymorphism in this receptor. Contrary to expectation, Fc gamma RIIINK from all donors irrespective of their NA allotype contained two bands (20 and 24 kDa) on SDS-PAGE after deglycosylation. In addition, those distinct epitopes on the extracellular domain of Fc gamma RIIIPMN found with mAb B73.1 and CLB gran 11 in association with the NA allotypic differences are expressed (or not expressed) on Fc gamma RIIINK independent of donor NA allotype. Fc gamma RIIIPMN and Fc gamma RIIINK differ at the protein level as they have different m.w. (glycosylated and deglycosylated), different epitopes in the extracellular domain (not attributable to tissue-specific glycosylation), and differential expression of the NA allelic protein polymorphism. Although the membrane anchor of Fc gamma RIIIPMN is a phosphatidylinositol-specific phospholipase C sensitive glycosyl-phosphatidylinositol linkage, Fc gamma RIIINK is insensitive to phosphatidylinositol-specific phospholipase C. However, a form of Fc gamma RIIINK is released from NK cells upon incubation at 37 degrees C. Thus, the basis for the two bands in Fc gamma RIIINK after N-linked deglycosylation is neither coexpression of two molecular isoforms with different membrane anchors nor an identifiable allelic polymorphism in m.w. restricted to Fc gamma RIIINK (p less than 10(-6)). The differences between the two receptors indicate that, independent of cell anchor type, PMN and mononuclear cells must have different molecular isoforms. The allelic variants, different isoforms, alternative anchor mechanisms and release processes provide for an extensive genetic and regulatory diversity in Fc gamma RIII function.

Differences among immune complexes: association of C1q in SLE immune complexes with renal disease.

Studies that made use of multiple assay systems demonstrated increased levels of immune complexes (IC) in patients with systemic lupus erythematosus (SLE), but no consistent correlations of IC concentration to patterns or activity of disease have been observed. Furthermore, consistent associations between qualitative differences in IC and disease manifestations have been elusive. IC interaction with erythrocytes and mononuclear phagocytic cells is another variable in SLE that may also mediate some of the biological effects of IC. The present report concerns studies of the composition of purified IC obtained from individuals with SLE and other rheumatic diseases; a 64,000 dalton component identified as the A-B subunit of C1q was detected in purified IC from 27 of 51 SLE patients (53%). The presence of this 64,000 dalton component was not related to either IC concentration or to the serum C1q level. However, the presence of the C1q component in isolated SLE IC did correlate with the presence of renal disease (p less than 0.02). These observations are interpreted relative to a recently described kinetic model of IC clearance.

Characterization of sequential immune complexes in infective endocarditis by Western blot analysis.

A patient with cutaneous vasculitis during infective endocarditis due to Lactobacillus casei was studied. Immune complexes (IC) were isolated from serum at the time of diagnosis and after 4 wk of therapy. Purification of IC used differential polyethylene glycol precipitation and competitive binding to staphylococcal protein A. In situ radioiodination of IC was performed, followed by SDS-polyacrylamide gel electrophoresis (PAGE). Anti-IC antisera were raised in rabbits by immunization with purified IC. IC were characterized by SDS-PAGE followed by electrophoretic transfer to nitrocellulose, incubation with antiserum and then with 125I protein A, and autoradiography. Although early and late IC differed quantitatively, there were no differentiating immunochemical features. Both IC contained a 60,000 dalton component that did not react with preimmune serum nor with anti-normal human serum. This component reacted with antiserum rendered specific for L. casei by affinity chromatography. The restricted antigen-antibody representation in IC contrasted with a wider panel of antibody activity in patient serum. The Western blot analysis proves to be an ideal method for the characterization of IC because of its sensitivity, dissociative capability, and preservation of immunoreactivity. IC isolated at a time removed from the original antigenic challenge may provide insight into the nature of the inciting antigen.

Reactive oxygen intermediates enhance Fc gamma receptor signaling and amplify phagocytic capacity.

Receptors for the Fc region of IgG (Fc gamma R) mediate internalization of opsonized particles by human neutrophils (PMN) and mononuclear phagocytes. Cross-linking of Fc gamma R leads to activation of protein tyrosine kinases and phosphorylation of immunoreceptor tyrosine-based activation motifs (ITAMs) within Fc gamma R subunits, both obligatory early signals for phagocytosis. Human PMN constitutively express two structurally distinct Fc gamma R, Fc gamma RIIa and Fc gamma RIIIb, and can be induced to express Fc gamma RI by IFN-gamma. We have previously shown that stimulation of PMN through Fc gamma RIIIb results in enhanced Fc gamma RIIa-mediated phagocytic activity that is inhibited by catalase. In the present study, we have tested the hypothesis that reactive oxygen intermediates (ROI) have the capacity to regulate Fc gamma R responses and defined a mechanism for this effect. We show that H2O2 augmented phagocytosis mediated by Fc gamma RIIa and Fc gamma RI in PMN and amplified receptor-triggered tyrosine phosphorylation of Fc gamma R-associated ITAMs and signaling elements. Generation of endogenous oxidants in PMN by cross-linking Fc gamma RIIIb similarly enhanced phosphorylation of Fc gamma RIIa and Syk, a tyrosine kinase required for phagocytic function, in a catalase-sensitive manner. Our results provide a mechanism for priming phagocytes for enhanced responses to receptor-driven effects. ROI generated in an inflammatory milieu may stimulate quiescent cells to rapidly increase the magnitude of their effector function. Indeed, human monocytes incubated in the presence of stimulated PMN showed oxidant-induced increases in Fc gamma RIIa-mediated phagocytosis. Definition of the role of oxidants as amplifiers of Fc gamma R signaling identifies a target for therapeutic intervention in immune complex-mediated tissue injury.

Fc gamma RIII expressed on cultured monocytes is a N-glycosylated transmembrane protein distinct from Fc gamma RIII expressed on natural killer cells.

Fc gamma RIII is a family of protein isoforms encoded by at least two distinct, yet highly homologous, genes. Fc gamma RIII on neutrophils is a glycosylphosphatidylinositol-linked protein with an allelic polymorphism (NA1/NA2) while Fc gamma RIII on NK cells (Fc gamma RIIINK) is an exclusively transmembrane protein without the NA polymorphism. The relationship of the isoform of Fc gamma RIII expressed on cultured monocytes (Fc gamma RIIIM phi) to these two forms, however, is unclear because some evidence suggests lowered expression of Fc gamma RIIIM phi in paroxysmal nocturnal hemoglobinuria (unlike Fc gamma RIIINK) and a unique deglycosylated m.w. for Fc gamma RIIIM phi. In this study we demonstrate that, as with Fc gamma RIIINK, Fc gamma RIIIM phi is resistant to the action of phosphatidylinositol-specific phospholipase C and is expressed at normal levels on affected (glycosylphosphatidylinositol-anchor negative) cultured monocytes from patients with paroxysmal nocturnal hemoglobinuria. Fc gamma RIIIM phi is also shed from the cell surface upon incubation at 37 degrees C. However, Fc gamma RIIIM phi and Fc gamma RIIINK have different m.w. as glycosylated proteins despite the same deglycosylated m.w. Thus, each cell type appears to express distinct glycoforms. These differences in glycosylation may influence the functional properties of the receptor.

Electrophoretic transfer blotting analysis of immune complexes in rheumatoid arthritis.

Immunochemical analysis of immune complexes (IC) derived from synovial fluid and serum of patients with rheumatoid arthritis (RA) was undertaken. IC were isolated by differential polyethylene glycol precipitation and competitive binding to staphylococcal Protein A. Anti-IC antisera were raised in rabbits by immunization with purified IC. IC were characterized by SDS-polyacrylamide gel electrophoresis (SDS-PAGE), electrophoretic transfer to nitrocellulose sheets, reaction with either anti-IC antiserum or monospecific antiserum against putative components, then developed with 125I-Protein A and autoradiography. Synovial fluid IC from both RA and control inflammatory joint diseases contained immunoglobulin light chains, and mu and gamma heavy chains. IC from RA synovial fluid also contained alpha chain, and Clq. Cross-over studies of IC and antisera revealed cross-reactive antigens in IC of RA synovial fluid and that remained unidentified were of the following relative molecular masses: 26,000; 39,000; 40,000; and 43,000. Studies using early and late serum samples to probe the blotted IC failed to disclose serological autoreactivity. This sensitive analytical technique provides a means for detailed description of the constituents of IC in inflammatory joint diseases.

Differential modulation of stimulatory and inhibitory Fc gamma receptors on human monocytes by Th1 and Th2 cytokines.

Immune complex-mediated inflammatory responses are initiated by Fc gamma R on phagocytes. We report in this study that an inhibitory receptor, Fc gamma RIIb2, is expressed on circulating human monocytes, and when co-cross-linked with stimulatory Fc gamma R it down-regulates effector function. Fc gamma RIIb2 expression is increased by IL-4 and decreased by IFN-gamma, in contrast to the activating receptor, Fc gamma RIIa, which is increased by IFN-gamma and decreased by IL-4. Thus, Th1 and Th2 cytokines differentially regulate the opposing Fc gamma R systems, altering the balance of activating and inhibiting Fc gamma R. The detection and cytokine modulation of Fc gamma RIIb2 in human myeloid cells provide evidence of a negative regulator of immune complex-mediated responses in human phagocytes and offer a new approach to limit Ab-triggered inflammation in autoimmune disease.

Anti-neutrophil cytoplasmic antibodies engage and activate human neutrophils via Fc gamma RIIa.

The presence of anti-neutrophil cytoplasmic Abs (ANCA) in many patients with systemic vasculitis suggests that ANCA may play a role in disease pathogenesis. Neutrophils from patients with Wegener's granulomatosis often express ANCA target Ags (myeloperoxidase (MPO) and proteinase 3 (PR3)) on their surface, making these intracellular primary granule enzymes accessible to these autoantibodies. Similarly, normal neutrophils can be induced to translocate MPO and PR3 to the cell surface in vitro, and we demonstrate that murine mAb ANCA IgG, but not IgM, binds to the ANCA target and engages the Fc gamma RIIa ligand-binding site on the surface of human neutrophils. In contrast to ANCA IgM, ANCA IgG also induces an oxidative burst in neutrophils (oxidation of dihydrorhodamine = 91 +/- 15 fluorescence units with anti-PR3 IgG vs 17 +/- 2 with anti-PR3 IgM, p < 0.001). Blockade of the ligand-binding site of Fc gamma RIIa with an antibinding site mAb Fab significantly reduces this ANCA IgG-triggered production of reactive oxygen species (p < 0.01). Similarly, human ANCA bind the ANCA target, engage Fc gamma RIIa, and induce an oxidative burst in neutrophils. The allelic phenotype of Fc gamma RIIa strongly influences the Fc gamma receptor engagement by ligand, and Fc gamma RIIa homozygous donors differ by more than threefold in the quantitative production of reactive oxygen intermediates (ROI) (p < 0.01). Thus, engagement of Fc gamma RIIa by the Fc region of ANCA is one mechanism by which these autoantibodies activate receptor-mediated signal transduction systems in human neutrophils to initiate programs of inflammation and tissue injury. Fc gamma receptor alleles may represent heritable disease risk factors influencing the magnitude of such a process.

Serial circulating immune complexes and mononuclear phagocyte system function in infective endocarditis.

Twenty patients with infective endocarditis were followed prospectively and all had elevated levels of circulating immune complexes (CICs) detected by staphylococcal binding assay. Mean CIC levels declined for the group as a whole (193 micrograms/ml +/- 24 to 100 +/- 17, p less than 0.05) and became undetectable in eight patients (47%) who were cured. Patients who died or had complicated courses had higher mean CIC levels at the start and finish (254 micrograms/ml +/- 24 and 145 +/- 37) of antibiotic therapy than patients with uncomplicated courses (178 micrograms/ml +/- 19 and 38 +/- 24), p less than 0.05. CIC levels did not decline significantly in patients with glomerulonephritis or arthritis, in contrast to patients without these features. Despite elevated CIC levels, 10 patients had enhanced mononuclear phagocyte system (MPS) function as assessed by Fc-dependent IgG-coated red blood cell clearance. These data suggest that CICs probably are pathogenic in endocarditis and may contribute to the development of arthritis and glomerulonephritis. Elevated CICs in infective endocarditis do not appear to be directly related to defective MPS function.

Carbohydrates on human Fc gamma receptors. Interdependence of the classical IgG and nonclassical lectin-binding sites on human Fc gamma RIII expressed on neutrophils.

To explore the molecular basis for the ability of aggregated IgG to block the phagocytosis by human polymorphonuclear leukocytes of Con A-opsonized E and of nonopsonized Escherichia coli with mannose-binding adhesins, we examined specific aspects of the glycoprotein structure of both the 40- to 43-kDa receptor for the Fc portion of IgG (Fc gamma RII) and the 50- to 78-kDa receptor for the Fc portion of IgG (Fc gamma RIIIPMN) from human polymorphonuclear leukocytes. Fc gamma RIIIPMN isolated by both mAb and ligand affinity chromatography, but not Fc gamma RII, binds Con A in Western blots. This binding is specifically inhibitable by alpha-methylmannoside. Digestion of Fc gamma RIIIPMN by recombinant endoglycosidase H, which is specific for high mannose-type (Con A-binding) oligosaccharides, alters the epitope recognized by mAb 3G8 in or near the IgG ligand-binding site of the receptor. Similarly, the ability of Fc gamma RIIIPMN to bind human IgG ligand is sensitive to endoglycosidase H digestion. Our data indicate that ligands other than the classical IgG opsonins can bind to human Fc gamma RIIIPMN per se through lectin-carbohydrate interactions. Furthermore, Fc gamma RIIIPMN contains a high mannose type oligosaccharide chain which contributes importantly to the integrity of the classical IgG ligand-binding site. Thus, specific glycosylations of the receptor are important for both classical and nonclassical engagement of Fc gamma RIII and may play a role in determining the properties of the ligand-binding site.

Occurrence of antibodies to Borrelia burgdorferi in patients with nonspirochetal subacute bacterial endocarditis.

OBJECTIVE: To determine the prevalence and specificity of antibodies to Borrelia burgdorferi in patients with nonspirochetal subacute bacterial endocarditis and assess whether increased levels of antibodies to B. burgdorferi were attributable to rheumatoid factor. DESIGN: Retrospective case-control study. SETTING: Urban referral center in an area devoid of infected ticks as a source of endocarditis sera. PATIENTS: Sera from 30 consecutive patients with culture-proven subacute endocarditis between 1979 and 1981 were compared with 30 control sera collected between 1989 and 1990. In addition, sera from 20 consecutive patients with rheumatoid arthritis who were positive for rheumatoid factor were collected between 1991 and 1992. Sera were compared with a convenience sample from 15 patients who met the criteria for Lyme disease. MEASUREMENTS: Antibodies to B. burgdorferi were assessed by enzyme-linked immunosorbent assay (ELISA) and immunoblot analysis. IgM rheumatoid factor was quantified using solid-phase radioimmunoassay or latex agglutination techniques. RESULTS: Thirteen of 30 patients with endocarditis (43%) compared with 3 of 30 normal controls (10%) had increased levels of antibodies to B. burgdorferi (P < 0.01). Of these 13 patients, only 1 had an immunoblot consistent with previous infection. The others had nonspecific immunoblots: 5 showed isolated 60-kd reactivity; 1 patient had isolated 41-kd reactivity; and 6 had no bands of reactivity. Immunoblots of the 3 controls with increased antibodies showed only isolated 41-kd reactivity. Thus, the specificity of the B. burgdorferi antibody test in patients with endocarditis was only 60% (95% CI, 42% to 78%), compared with 90% (CI, 79% to 100%) in controls. No correlation was noted between IgM rheumatoid factor and antibodies to B. burgdorferi in patients with endocarditis (r = 0.2; P > 0.2). Only 1 of 20 patients with rheumatoid arthritis without known bacterial infections had antibodies to B. burgdorferi. CONCLUSIONS: Although a positive ELISA test for B. burgdorferi may be a "true positive," a positive serologic test alone does not ensure that the clinical problem is due to Lyme borreliosis. Cross-reactive antibodies to shared epitopes between B. burgdorferi and the endocarditis organism may account for the high false-positive results.