Common epitopes in Clq and collagen type II

Pathogenesis of renal disease in systemic lupus erythematosus--the role of autoantibodies and lymphocytes subset abnormalities

Lupus nephritis (LN) is a common and severe organ manifestation of systemic lupus erythematosus (SLE), and is associated with significant patient morbidity and mortality. Autoantibodies and aberrations in lymphocyte subsets have putative roles in the pathogenesis of SLE and LN, and might reflect disease activity and are amenable to immunosuppressive treatments. Anti-DNA is one of the well-studied autoantibodies, which correlates with disease activity and has direct nephritogenic effects on resident renal cells and various glomerular components. Other important autoantibodies in the pathogenesis of LN include anti-C1q, anti-α-actinin and anti-nucleosome antibodies. Changes in naive and memory B cells and plasma cells have been observed in SLE and LN patients. These B cell subsets exert diverse effects during pathogenesis of LN such as production of autoantibodies, secretion of proinflammatory and anti-inflammatory cytokines and presentation of auto-antigens to effector cells. Aberration of T lymphocytes, especially the T-helper subsets, is also highly pertinent in the development of LN. In this context, important T helper subsets include Th1, Th2, Th9, Th17, T_Reg and follicular T-helper cells. The growing knowledge on these autoantibodies and lymphocyte subset abnormalities will enhance our understanding of SLE and LN, and hence help devise better strategies for disease monitoring and treatment.

Review: Lupus nephritis: pathologic features, epidemiology and a guide to therapeutic decisions

Systemic lupus erythematosus may present with renal manifestations that frequently are difficult to categorize and lupus nephritis is an important predictor of poor outcome. The type and spectrum of renal injury may remain undiagnosed until full-blown nephritic and/or nephrotic syndrome appear with increased risk of end-stage renal disease. These abnormalities occur within the first few years after the diagnosis of lupus is made on clinical grounds and with the support of laboratory tests in high risk patients. An early renal biopsy is helpful in patients with an abnormal urinalysis and/or reduced glomerular filtration rate and the results form the basis for therapeutic decisions. The biopsy also provides vital prognostic information based on histological categorization of different types of lupus nephritis, the degree of activity, chronicity and the immunopathogenesis. In the current armamentarium, the use of cyclophosphamide and azathioprine and recently mycophenolate mofetil, reduce morbidity and maintenance therapies reduce the risk of end-stage renal disease. Clinical trials underway promise new, effective and safe immunosuppressive regimens for the treatment of proliferative lupus nephritis.

Humoral autoreactivity directed against surfactant protein-A (SP-A) in rheumatoid arthritis synovial fluids

SP-A is found principally in the lung, and has been associated with lamellar bodies also found in the synovial joint. Both SP-A and C1q contain collagen-like regions, and SP-A and C1q have some structural similarities, both having a globular head region and a collagen-like tail. Here we are able to show that (i) autoreactivity to SP-A, as expressed by IgG and IgM autoantibodies, is present in synovial fluid (SF) isolated from patients with rheumatoid arthritis (RA); (ii) in absorption experiments only a limited degree of cross-reactivity between autoantibodies reactive with C1q and SP-A is observed; (iii) there is no cross-reactivity between autoantibodies reactive with type II collagen (CII) and those reactive with SP-A or C1q; (iv) autoantibodies react with polymeric (dimers and larger) SP-A, but not with monomeric SP-A subunits, indicating that a degree of quaternary structure is required for antibody binding. Unlike CII, which not accessible in the normal joint, both SP-A and C1q are available within the SF in patients with RA and may therefore provide antigens driving an autoimmune response directed against collagen-like structures.

Correlation between anti-C1q and immune conglutinin levels, but not between levels of antibodies to the structurally related autoantigens C1q and type II collagen in SLE or RA

The simultaneous appearance of autoantibodies with either a functional or structural relationship to anti-C1q antibodies (anti-C1q) was investigated in 39 systemic lupus erythematosus (SLE) patients and in 28 rheumatoid arthritis (RA) patients, in both cross-sectional and longitudinal design. Levels of anti-C1q showed an isotype-specific correlation to levels of immune con-glutinin (IK) in SLE patients, whereas no correlation was evident to levels of antibodies to the structurally related antigen type II collagen (anti-CII) in SLE or RA patients. IgG anti-C1q levels correlated with serum levels of the terminal complement complex (sC5b-9) in SLE patients. In two longi-tudinally followed patients, the IK response preceded the anti-C1q response. Possibilities for regulation of the humoral anti-complement response are discussed.

The collagen-like component of the complement system, C1q, is recognized by 7 S autoantibodies and is functionally impaired in synovial fluids of patients with rheumatoid arthritis

Cross-reactivity between type II collagen (CII) and C1q, the collagen-like subunit of the first component of complement, has been demonstrated in synovial fluid (SF) from rheumatoid arthritis (RA) patients. Many authors have studied autoimmunity to CII in RA, but little work has been done on autoimmunity to C1q in RA. In the data presented here, we have been able to show that in addition to native C1q, an altered form of C1q is present in SF from RA patients. Furthermore, a low molecular weight form of C1q is present in RA SF, although its role, if any, in the pathogenesis of RA is unclear. The presence in these RA SF of C1q-specific antibodies (IgG and IgM) has been studied and we have partially characterized the antibody moieties involved. As well as binding to C1q and fragments representing the collagen-tails from C1q, 7 S IgG autoantibodies against C1q also bind to a C1q molecule altered in vitro by incubation with reactive oxygen species and to the non-apeptide KGEQGEPGA (representing residues 26-34 from the C1q A-chain), which has previously been shown to suppress the onset of CII-induced arthritis in an animal model.

Complement components C1q, C1r/C1s, and C1INH in rheumatoid arthritis. Correlation of in situ hybridization and northern blot results with function and protein concentration in synovium and primary cell cultures

OBJECTIVE

To analyze the synovial site and the cell types expressing C1q, C1r/C1s, and C1-esterase inhibitor (C1INH) and to characterize newly synthesized C1q in patients with rheumatoid arthritis (RA).

METHODS

Tissue and primary cell cultures of synovium from RA patients were analyzed for C1q, C1r/C1s, and C1INH by Northern blotting, in situ hybridization, and pulse-chase experiments for C1q.

RESULTS

The de novo synthesis of C1q, C1r/C1s, and C1INH in synovium and primary cell cultures was proven by Northern blot and by antigenic and functional analysis. In in situ hybridization experiments, the synovial lining cell layer was identified as the site of C1q, C1r, and C1INH expression. In contrast, immunohistologic analysis showed that C1q, C1s, and C1INH proteins were present in a thin film covering the synovial lining cells. In situ hybridization performed on primary cell cultures provided evidence that only macrophages were able to express C1q, whereas fibroblasts and stellate cells synthesized C1r.

CONCLUSION

The synovium is important for the synthesis and secretion of C1q and C1r/C1s, as well as the control protein C1INH, which supports the idea of a locally occurring inflammatory process in RA patients.

Local versus systemic immunoreactivity to collagen and the collage-like region of C1q in rheumatoid arthritis and SLE

In this paper we discuss recent data concerning immunoreactivity on the one hand to the cartilage derived molecule type II collagen (CII) and on the other hand to the complement factor C1q, in rheumatoid arthritis (RA) and systemic lupus erythematosus (SLE). By using the ELISPOT method we have been able to show that production of antibodies to CII and to C1q is common in inflamed joints but not in the periphery of RA-patients, whereas production to C1q is commonly seen in the circulation of active SLE patients. Previously an immunological cross-reactivity has been described between these two proteins. The possibility of in vivo cross-reactivity between these two molecules might have consequences for the different disease manifestations in RA and SLE. This paper also stresses the importance of single cell analysis of antibody production when evaluating local antibody production in different body compartments and also when the aim is to assay changes in antibody production over time.

Short-term kinetics of the humoral anti-C1q response in SLE using the ELISPOT method: fast decline in production in response to steroids

Twenty four systemic lupus erythaematosus patients and 17 patients with other diagnoses were investigated regarding the presence of cells producing C1q reactive antibodies in peripheral blood mononuclear cells using the ELISPOT technique. These results were then compared with parallel serum levels of C1q reactive antibodies. Current production of anti-C1q was almost entirely confined to the systemic lupus erythaematosus group. Longitudinal analysis of anti-C1q ELISPOT positive patients showed rapid changes in the number of anti-C1q producing cells, but only slowly changing serum levels of the corresponding antibodies in response to glucocorticoids. In one systemic lupus erythaematosus patient prednisolone treatment had a selective effect on this autoantibody production, as the production of anti-C1q spot forming cells rapidly dropped to zero, at the same time as the number of total spot-forming cells showed only less change. In another patient, self-limiting connective tissue disease was associated with temporal occurrence of IgM anti-C1q. We believe, from these data, that the ELISPOT method for determination of current antibody production may be of particular value in longitudinal evaluation of disease course and therapeutic effects in systemic lupus erythaematosus and other rheumatic diseases.

Inhibitor of C1q secretion suppresses the macrophage response to lipid A for nitric oxide but not for TNF production: evidence for a role of C1q in autocrine binding of TNF

Studies were designed to further define the modulatory role of complement subcomponent C1q in macrophage activation by Lipid A to mediate production of TNF and cytotoxic nitric oxide (NO). Pretreatment of macrophages for 24 h with 2.5 mM 3,4,dehydro-D,L-proline (DHP), an inhibitor of C1q secretion, suppressed their response to Lipid A activation for cytotoxicity of P815 tumor targets which correlated with a corresponding decrease in NO production. In contrast, DHP-pretreated macrophages displayed an increase in the release of TNF in response to Lipid A as compared to untreated controls. Time kinetic studies indicated that DHP-pretreated macrophages produced higher sustained levels of TNF activity during 1 to 24 h culture with Lipid A than did untreated control macrophages. This was confirmed by increased TNF mRNA expression in response to Lipid A by DHP-treated cells. DHP-pretreated macrophages had reduced levels of cell surface C1q as determined by cytofluorometric analysis of the binding of FITC-labeled anti-C1q, F(ab')2. Macrophages were also found to have reduced binding capacity for phycoerythrin-labeled rTNF (PE-TNF) by cytofluorometric analysis following DHP treatment. Exposure of DHP-pretreated macrophage to soluble C1q at 4 degrees C restored their reduced binding of PE-TNF. C1q was confirmed to bind to macrophages at 4 degrees C as detected by FITC anti-C1q, F(ab')2 and such C1q binding promoted a corresponding increased binding of PE-TNF. Macrophages which were plated over immobilized C1q were also markedly enhanced in their binding of PE-TNF probe. Our results indicate that the inhibition of macrophage secretion of C1q by DHP pretreatment, was accompanied by an increased TNF mRNA expression and release with a decrease in NO generation following Lipid A activation. Since TNF binding to DHP-treated macrophages was reconstituted by the binding of exogenous C1q to the cells, it appears that C1q may be involved in the modulation of autocrine binding of TNF for subsequent generation of cytotoxic NO.

Autoimmune recognition of cartilage collagens

Recent advances in basic research on the immune system and molecular biology of cartilage components have greatly increased our understanding of the role of autoimmunity in inflammatory diseases affecting joints, particularly rheumatoid arthritis. Many of these diseases are common and their complex pathogenesis probably involves a large number of genes polymorphic in the population as well as environmental factors. Characteristic features of inflammatory arthritis include expansion of the synovial tissue into a pannus containing lymphocytes and macrophages, autoimmune reactions against cartilage antigens, and erosion of cartilage. Since hyaline cartilage of the articular surfaces is the only structure within the joint known to contain joint-specific antigens this tissue is the prime suspect as the target of the autoimmune This review will first present the capacity of the immune system to discriminate between self and non-self structures, and then summarize our current understanding of the structures of cartilage collagens. Subsequently we will discuss how the immune system normally interacts with cartilage and how such interactions can lead to arthritis. We propose that collagen-induced arthritis (CIA) is valuable for understanding the autoimmune recognition of cartilage collagen which precedes the outbreak of arthritis and may perpetuate its chronicity, and serves as an animal model of rheumatoid arthritis.

Western blot analysis of human IgG reactive with the collagenous portion of C1q: evidence of distinct binding specificities

An enzyme-linked immunosorbent assay (ELISA) with purified collagenous C1q fragments in the solid phase was used for detection of C1q-specific immunoglobulins in the sera of twelve patients with systemic lupus erythematosus (SLE) or the SLE-like disease hypocomplementemic urticarial vasculitis syndrome (HUVS). By clinical criteria, four patients had SLE, and three HUVS. Five patients had overlap syndromes. All patients demonstrated high concentrations of C1q-specific IgG and markedly low concentrations of circulating C1q. Detection of C1q-specific IgG in SLE sera was facilitated by employment of saturating concentrations of collagenous C1q fragments in the solid-phase ELISA. When added to SLE serum, immune complex-fixed C1q inhibited binding of IgG to the C1q fragments, whereas addition of C1q alone had limited inhibitory effects. Under similar conditions, using approximately equimolar amounts of C1q relative to solid-phase C1q fragments, no ELISA inhibition was obtained after addition of C1q or immune complex-fixed C1q to a HUVS serum. Even in large excess, purified C1q did not inhibit binding of HUVS-IgG to solid-phase C1q fragments. Thus, possible interactions between HUVS-IgG and native Clq are probably of low affinity. By Western blot analysis, IgG reactive with the B and C chains of C1q was found in the eight patients with evidence of HUVS, five of whom also showed IgG binding to C'-C' and A'-B' dimers of collagenous C1q fragments. Sera from SLE patients were negative by Western blot analysis. It seems likely that C1q-specific IgG in SLE primarily recognizes assembled C1q molecules or collagenous C1q fragments expressing conformational epitopes of bound C1q. Interestingly, patients with evidence of HUVS fairly consistently had zymogen (C1r-C1s)2 complexes in their serum, while patients with SLE showed high concentrations of complexes containing Cl inhibitor, C1r and C1s. Different binding specificities of C1q-reactive IgG could be of importance with regard to pathogenetic mechanisms in SLE and HUVS. There was no correlation between findings of C1q-specific IgG and a variety of autoantibodies associated with SLE and SLE-like disease.

Modulation of type II collagen-induced arthritis in DBA/1 mice by intravenous application of a peptide from the C1q-A chain

In this report we are able to show that intravenous (i.v.) application (day 0) of a nonapeptide (residues 26-34) from the human C1q A-chain (designated peptide A-C1q) prior to intradermal (i.d.) administration of chicken type II collagen (CII) in arthritis-susceptible DBA/1 mice (H2q), leads to abrogation of polymorphonuclear neutrophil (PMN) invasion into the joints. This nonapeptide exhibits epitope characteristics and high homology to residues 137-147 of CB11 (a cyanogen bromide fragment of chicken CII, known to contain both arthritis inducing and suppressing determinants). Arthritis index was lowest in animals pretreated i.v. with CII (as internal control), though animals pretreated i.v. with peptide K (residues 137-147 with an additional glycine residue from CB11) or peptide A-C1q exhibited comparative arthritic indices. Only in the arthritis-positive control group (day 0: PBS i.v.) did i.d. application of CII lead to invasion of PMN into the synovial layer and the joint space. Analysis of antibody (Ab) responses at day 48 after i.v. immunization (day 0) and CII challenge (day 7) revealed IgE-Abs to native CII and also to native C1q. IgG titers to CII were highest in animals pretreated with peptide A-C1q. Abs from this group, exhibiting activity to peptide A-C1q (immunizing antigen), were of mainly IgG1 and IgG3 isotypes. Evaluation of the immune response following i.v. application of peptide A-C1q or CII, prior to i.d. CII administration, in DBA/1 mice, revealed IgM responses to peptide A-C1q and peptide K, but not to CII. Intravenous application of peptide A-C1q led to generation of IgG3-Abs reacting only with peptide A-C1q and peptide K, but not with native CII. Additionally, i.v. application of peptide A-C1q elicited IgG responses to a pentapeptide, resembling amino acid residues 26-30 (K-G-E-Q-G) of the C1q A-chain. This five residue antigenic determinant is present in peptide K, in chicken and human CII as well as in human C1q. No specific IgE response to any of the antigens tested could be detected. Since a peptide from the C1q A-chain is both capable of eliciting immune responses and modulating CII-induced arthritis in mice, we postulate that the collagen-like complement component C1q is involved in the development of CII-induced inflammatory arthritic lesions, and may represent, in vivo, the early antigen responsible for inducing anticollagen antibodies prior to CII in hyaline cartilage becoming available as antigen.

Rheumatoid arthritis

The immunopathogenesis of rheumatoid arthritis is discussed in two ways. First, we consider the major question of whether T cells are likely to drive the disease. Second--and assuming T cells to be important--we discuss available data on the components of the trimolecular complex (major histocompatibility complex class II-antigen-T-cell receptor), which are possibly involved in the disease. Our two main points are that the most important questions concerning the pathogenesis of rheumatoid arthritis require answers from immunointervention in patients, and that animal experiments can be increasingly used in interpreting current experiments in humans.

Multiple epitopes on cartilage type II collagen are accessible for antibody binding in vivo

Monoclonal mouse antibodies specific for the major epitopes on mouse type II collagen (CII) were biotinylated and injected into neonatal and adult mice. Anti-CII antibodies, specific for four different epitopes on the CII molecule, could be shown to bind specifically to joint surfaces in the paws of 2-day-old syngeneic DBA/1 mice after an intraperitoneal injection of 100 micrograms of biotinylated antibody. The anti-CII antibodies did not bind to cartilage from DBA/1 mice in vitro, unless the sections were pretreated with hyaluronidase or the specimens decalcified prior to freezing, showing that the epitopes are accessible in vivo but not in vitro. By analyzing the in vivo binding capacity for a number of monoclonal anti-CII antibodies which represented different IgG subclasses, it could be demonstrated that binding to the same epitopes occurred independent of IgG subclass. However, one epitope (denoted "B1") was only weakly detected, possibly due to the fact that the antibody used (CIIB1) crossreacts with type I collagen and C1q. Monoclonal anti-CII antibodies, injected into neonates or adult mice, bound specifically to most, but not all, tissues containing CII; including hyaline joint cartilage, fibrous sternal and costal cartilage, tracheal cartilage and fibrous cartilage in the spine but not to CII-containing structures in the eye. The finding that CII, while present in cartilage, is accessible for antibody binding in vivo may have important implications for the availability of CII for the immune system and for the understanding of the development of pathological autoimmunity leading to collagen-induced arthritis in mice.

Arthritis in DBA/1 mice induced with passively transferred type II collagen immune serum. Immunohistopathology and serum levels of anti-type II collagen auto-antibodies

Arthritis was induced in DBA/1 mice by passive transfer of syngeneic anti-type II collagen (CII) serum concentrate. After transfer of serum containing 0.2 or 0.5 mg anti-CII auto-antibodies the first clinical signs of arthritis appeared 48 h after injection. Severe clinical arthritis was detected 96 h after injection. Immunohistochemical analyses of joints 48 h after serum injection revealed synovial foci in intercarpal and metacarpophalangeal joints of macrophage-like cells, expressing C3bi-receptors and major histocompatibility complex class II molecules, and infiltration of few CD4+ lymphocytes. Later (96 h after injection), the inflamed synovia were dominated by C3bi-receptor+ polymorphonuclear cells. In contrast to conventionally induced collagen arthritis (CIA), the inflammatory infiltrates, filling joint spaces and synovial tissue, were extensively dominated by polymorphonuclear cells, whereas macrophage-like cells expressing class II molecules and a few T cells were seen only in the periphery of the developing pannus. The anti-CII serum induced arthritis may be used as a model for studies of humoral mediated mechanisms operating in conventionally induced CIA as well as in rheumatoid arthritis.