Autoantibodies against C1q: view on clinical relevance and pathogenic roles

Immune responses to autoantigens are quite common and autoimmunity is considered to be a physiological part of the immune system. Autoantibodies are an inherent property of the antibody repertoire of many healthy individuals and are therefore referred to as natural autoantibodies [1,2]. The numerous kinds of autoantibodies fall into two main categories, depending on whether organ-specific or non-organ-specific autoantigens are involved. Non-organ-specific antigens mostly occur in nucleated cells, such as DNA, or are found among circulating plasma proteins, such as coagulant proteins and the Fc portion of IgG. Natural autoantibodies have been proposed to be involved in the clearance of degradation products that are formed during cell metabolism. In this respect natural autoantibodies are also referred to as housekeeping antibodies [3]. Autoantibodies may also be associated with disease states but do not necessarily play a role in the pathogenesis of such diseases. The presence of autoantibodies may for example be secondary to the production of tissue damage, or independent pathogenic factors may directly induce both the disease and the presence of autoantibodies. In 1984 autoantibodies to C1q (C1qAb) were reported to be present in serum of patients with systemic lupus erythematosus (SLE) [4]. The recognition that C1q may serve as a non-organ-specific autoantigen has attracted a growing number of investigators. This study discusses the knowledge of C1q as autoantigen by reviewing the epidemiology, disease associations, and pathophysiology of C1qAb. Activation of the complement system is the first step in the prevention of damage by immune complexes. Initiation of complement activation occurs through three pathways: the classical, the alternative, and the lectin pathway. The classical pathway of the complement system is considered to be the most important pathway in immune complex clearance. This pathway may be activated by IgM- and IgG-containing immune complexes after binding of C1q [5]. C1q is a subcomponent of the first component (C1) of the classical pathway. It is a large highly cationic glycoprotein with a molecular weight of 410 kD. C1q consists of six copies each of three polypeptide chains, A, B, and C. The A, B, and C chains are rich in hydroxylated amino acids and are linked together by disulphide bonds into dimers [6] Together these dimers form a triple helix structure which resembles collagen. Towards the N-terminal end of C1q the triple helices lay parallel to each other and towards the C-terminal end they diverge. The N-terminal end is called the collagen-like region which is linked by the connecting strands to the C-terminal end, which is called the globular heads region. The macromolecular structure of C1q is said to resemble a bunch of tulips [5–7]. The function of C1q is directly related to its structure. Binding of Fc regions of immunoglobulins to the globular head portions of C1q induces distortion of the connecting strand which changes the conformation of the collagen-like region [8,9]. The dynamic equilibrium between C1q and the other subcomponents of C1, C1r and C1s, subsequently shifts and induces further activation of the cascade of proteins composing the classical pathway. This results in the prevention of lattice formation of immune complexes and ensures their clearance from the circulation by the mononuclear phagocyte system [10]. Although the recognition protein of the lectin pathway, mannose-binding lectin, is structurally related to C1q, it is not known to be involved in immune complex clearance mechanisms [11]. To summarize, activation of C1 by binding of immune complexes to C1q is a prerequisite for immune complex clearance. Since SLE is considered to be the prototype of immune complex diseases in man, a large variety of immune complex assays has been employed to investigate possible pathogenic roles of circulating immune complexes and to relate their titres to the presence of manifestations of the disease. The solid-phase C1q binding assay is one of the most frequently used assays for both purposes [12]. This radioimmunoassay is based on the binding of immune complexes to C1q, which is fixed to a solid phase. Studies in the early 1970s on the nature of proteins bound by solid-phase C1q already revealed that apart from immune complexes, ‘unidentified low molecular weight (7S) C1q reactants’ were also bound [13]. Similar observations were later made by a number of other investigators [14–17]. At that time it was already noted that ‘7S C1q precipitins’ were especially found in serum of patients with lupus nephritis. The actual proof of the antibody nature of ‘7S C1q precipitins’ was provided in the 1980s [4]. Instead of binding of the Fc part of IgG as part of an immune complex, monomeric IgG was demonstrated to be able to bind to C1q through its antigen-binding region. Additional experiments revealed that isolated F(ab′)2 portions of IgG in serum of SLE patients reacted with epitopes on the collagen-like region of C1q [4,18–20]. In addition, this binding was shown to be of high affinity and not affected by the presence of fluid-phase C1q. Subsequent studies also demonstrated the presence of IgA C1qAb with similar characteristics in serum of patients with rheumatoid vasculitis [21]. The solid-phase C1q binding assay is frequently modified into an ELISA for the measurement of C1qAb. Since C1qAb epitopes are located on the collagen-like region of C1q, either whole C1q or the collagen-like region of C1q has been used for C1qAb measurement [18,22,23]. By analogy, true measurement of circulating immune complexes in the presence of C1qAb has been exerted using only the globular head region of C1q [24]. One pitfall in the measurement of C1qAb is possible interference by binding of immune complexes in assays using whole C1q. Such binding may be abrogated by increasing the ionic strength in the incubation media used above the physiological level. Binding of immune complexes was demonstrated to be completely inhibited when the sodium chloride concentration used exceeded 0.15 m [21]. A second pitfall in IgG C1qAb measurement is the binding of immune complexes containing small native dsDNA fragments to C1q. Native dsDNA was demonstrated to be able to bind to the collagen-like region of C1q [25]. False-positive results in the detection of IgG C1qAb might ensue by the binding of IgG anti-dsDNA as part of an immune complex to C1q [26]. However, the binding of dsDNA to the collagen-like region of C1q was also shown to be prevented by adjustment of the ionic strength of the incubation media used above the physiological level [27,28]. In contrast, binding of C1qAb was not affected by this procedure and incubation of SLE serum samples with DNase did not influence the results of the measurement of C1qAb titres [4,17,18,20,25]. These observations imply that assays used for C1qAb measurement should be performed under high ionic strength conditions to prevent false-positive results by (DNA-containing) immune complexes. A third pitfall in the measurement of IgG C1qAb may be recognition of C1q by anti-collagen antibodies. Proteolysis by collagenase or exposure to oxidation of C1q may reveal epitopes on the collagen-like region of C1q that also occur on collagen type II [29,30]. IgG autoantibodies to collagen have been demonstrated to be present in serum of SLE patients [31]. In a longitudinal comparison of C1qAb and collagen type II autoantibody titres in the same population of SLE patients, separate disease associations for each antibody were found [32]. The results of this study indicate that the assays used for the detection of C1qAb and collagen type II autoantibodies are specific and therefore apparently detect different antibodies. In addition, the use of native C1q or the collagen-like region of C1q under standardized conditions should exclude the possibility of proteolysis by collagenase and the induction of cross-reactivity. Further evidence for the specificity of C1qAb was obtained in experiments indicating that C1qAb do not cross-react with mannose-binding lectin or other collectins [33]. It is therefore unlikely that cross-reactive epitopes on collagen type II or mannose-binding lectin interfere in standardized assays used for the measurement of C1qAb. Similar to previous investigations concerning other autoantibodies, the occurrence of C1qAb as natural autoantibody in the general population has been investigated [34]. Titres of IgG and IgA C1qAb were found to increase at older ages. This phenomenon was most pronounced over the age of 50 years. In contrast, the highest IgG C1qAb titres were found in the group of SLE patients between 20 and 39 years old. During the last decade evidence has been provided for the presence of IgG C1qAb in an increasing number of autoimmune and renal diseases in adult and paediatric medicine (Table 1) [22,23,35–40]. Recently the occurrence of IgA C1qAb in serum of patients with IgA nephropathy was reported, though initial observations could not find proof for their existence in this common form of glomerulonephritis [35,39]. The distribution of IgG C1qAb subclasses in patients with SLE and membranoproliferative glomerulonephritis (MPGN) was also investigated. Initial observations showed that production of IgG C1qAb in SLE patients was not restricted to any IgG subclass and was comparable to the distribution found in normal sera [40]. However, subsequent reports demonstrated that restriction towards IgG2 and IgG3 in SLE patients and towards IgG3 in patients with MPGN does occur [41–43]. IgA C1qAb production involved IgA1 in all patients and IgA2 in about half the patients investigated [42]. The combined studies seem to indicate that the mere occurrence of C1qAb in any individual has no pathological significance or diagnostic importance. The occurrence of IgG autoantibodies against C1q (C1qAb) in adult and paediatric autoimmune and renal diseases* Data adapted from [22,23,35–40]. The occurrence of IgG autoantibodies against C1q (C1qAb) in adult and paediatric autoimmune and renal diseases* Data adapted from [22,23,35–40]. The occurrence of C1qAb in certain diseases may be associated with specific manifestations. Such associations may provide a practical tool for daily patient management by the serial measurement of C1qAb titres. The first cross-sectional study of associations between C1qAb titres and clinical variables of disease activity in SLE patients demonstrated that the highest C1qAb titres were found in patients with active lupus nephritis (Fig. 1) [38]. Subsequent prospective studies demonstrated that rises in C1qAb titres may predict ensuing relapses of lupus nephritis [32,44,45]. Manifestations of diffuse proliferative lupus nephritis especially were found to be associated with C1qAb. In a comparison between titres of anti-dsDNA antibodies and C1qAb, both autoantibodies were equally effective in predicting renal relapses of SLE. However, anti-dsDNA antibody titre increments were also followed by extrarenal disease manifestations, whereas this was not established for C1qAb. For C1qAb, renal relapses hardly ever occurred without preceding C1qAb titre rises and significant C1qAb titre rises were followed by a renal relapse in half the cases [44]. In a recently published longitudinal comparison a solid-phase enzyme-linked immunospot (ELISPOT) assay was employed and similar disease associations were observed [46]. Even if C1qAb titres might not always vary according to changes of disease activity in individual patients, their presence clearly is a marker of classical pathway activation [38,47]. The combined studies on the value of serial measurement of C1qAb titres in SLE patients indicate that changing C1qAb titres may become an important parameter for the guidance of immunosuppressive therapy in these patients. C1qAb titres (mean + s.d.) measured in serum of 80 healthy control individuals, 65 patients with systemic lupus erythematosus (SLE) without lupus nephritis (nephritis−), and 20 patients with lupus nephritis (nephritis+). *Significantly higher than titres in SLE patients without nephritis and in healthy control individuals. C1qAb titres (mean + s.d.) measured in serum of 80 healthy control individuals, 65 patients with systemic lupus erythematosus (SLE) without lupus nephritis (nephritis−), and 20 patients with lupus nephritis (nephritis+). *Significantly higher than titres in SLE patients without nephritis and in healthy control individuals. C1qAb may play a role in the pathogenesis of diseases by several mechanisms. Circumstantial evidence for possible pathogenic roles for C1qAb may be provided by the knowledge that important diseases that C1qAb are associated with, such as SLE and MPGN, are characterized by the presence of circulating immune complexes. Interaction between the complement system and immune complexes occurs through binding to C1q. C1q, as antigen for C1qAb, is only recognized when it is bound to a solid phase [19,20]. This phenomenon may be explained by the exposure of previously hidden epitopes on the collagen-like region C1q when bound to a solid phase [48,49]. This finding has implications for our understanding of possible pathogenic roles of C1qAb. Likely solid-phase situations in which C1q could occur are when bound to immune complexes or to tissue structures. C1qAb may contribute to the pathogenesis of immune complex disease by participating in immune complex formation. Immune complexes may be formed either in the circulation or locally in tissues. C1q molecules bound by immunoglobulins that are part of an immune complex are potential antigens for C1qAb. C1qAb may amplify the formation of immune complexes by increasing immune complex size. Among other factors, the size of immune complexes determines whether lattices are formed [50]. If subsequent trapping of circulating immune complexes occurs inflammatory reactions, such as the development of glomerulonephritis and vasculitis, may ensue. Circulating C1qAb may also bind to C1q that has already deposited in tissues, leading to local or in situ immune complex formation. The concept of local immune complex formation, when antibodies bind to endogenous or planted exogenous antigens, is generally considered to play an important role in the pathogenesis of glomerulonephritides. Especially preferential binding of antigens to the glomerular basement membrane (GBM) on the basis of charge–charge interactions may play a role in this type of immune complex formation. C1q, as a highly cationic protein, may be seen as an attractive candidate to bind to the negatively charged GBM. Subsequently, immune complexes may be locally formed in the presence of C1qAb. Like other immune complexes, this type of immune complex also potentially possesses the capacity to induce inflammation by activation of the complement system. Several observations provide evidence for the involvement of C1qAb in the systemic or local formation of immune complexes. These include the finding that especially large C1q-containing circulating immune complexes have been shown to be associated with the development with lupus nephritis [15,51]. In rheumatoid vasculitis, another immune complex-mediated autoimmune disease, C1qAb were also demonstrated to contribute to the formation of circulating immune complexes [21]. Immunofluorescence studies of renal biopsies of patients with SLE invariably demonstrated the presence of immunoglobulins and components of the classical pathway of the complement system in glomeruli [52]. In addition, the occurrence of significant increases of C1qAb titres preceding the development of glomerulonephritis in SLE patients provides further support for a pathogenic role of C1qAb in the development of nephritis. Several experimental in vivo studies further illustrate the importance of C1qAb for the development of nephritis. Two investigators were able to demonstrate that the administration of human C1q to rodents resulted in C1q deposition in glomeruli. After subsequent infusion of C1qAb, immune complexes were formed and glomerulonephritis developed [53,54]. Of great importance is a recent human study on eluates of kidneys obtained from SLE patients [55]. Antibodies directed against the collagen-like region of C1q were recovered from glomeruli of 4/5 patients suffering from diffuse proliferative lupus nephritis. C1qAb may influence the physiological role of C1q after binding of C1q to an immune complex. Binding of C1qAb to the collagen-like region of C1q may disturb the dynamic equilibrium with the other subcomponents of C1 and interfere with activation of C1. Decreased C1 activation could functionally resemble a C1q-deficient state and result in immune complex precipitation in the tissues. Alternatively, C1qAb may also stimulate C1 activation after binding to C1q. In vitro studies showed that C1 could be activated by binding of MoAbs against epitopes on the collagen-like region of C1q [56]. The in vivo association between C1qAb titres and hypocomplementaemia may result from such stimulation [38,57]. However, in vitro and in vivo studies on activation of the classical pathway in the presence of C1qAb could not demonstrate any influence [58]. We conclude that at present there is no evidence that C1qAb directly influence C1 activation. The data presently available suggest that C1qAb may exert a role in the pathogenesis of immune complex diseases by contributing to the deposition or formation of immune complexes in glomeruli, which results in local inflammation and the development of glomerulonephritis. C1qAb are natural autoantibodies that are strongly associated with immune complex diseases, most prominently with hypocomplementaemic urticarial vasculitis syndrome and diffuse proliferative lupus nephritis. Although the presence of C1qAb has not been proved to be diagnostic for any disease, the measurement of C1qAb titres over time may become a valuable tool for the clinical management of SLE patients. Recent studies indicate that C1qAb contribute to the formation of immune complexes and therefore may play an important role in the pathogenesis of lupus nephritis. This work was done in part in the framework of the Biomed-2 Program BMH4-CT96-1005, the role of complement in susceptibility to infection and chronic disease.