Lupus autoantibodies 101: one size does not fit all; however, specificity influences pathogenicity

Multiple organ involvement is typical in lupus patients; however, variation among individuals is the rule and, in many, different organ systems are involved over their lifetime. For example, a young woman can present initially with arthritis and skin rash and later in life develop nephritis and pleuritis, whereas another woman's disease may consist of relapsing bouts of leucopenia, anaemia and pericarditis. The explanation for these variations is not understood fully. In general, breakdown in immunological tolerance leads to the activation of autoreactive B and T cells that through either their secretory products (e.g. autoantibodies, cytokines) or/and direct infiltration (e.g. T cells or B cells) initiate disease [1]. The nature and severity of inflammation is determined both by the character and extent of the invasion, along with the systemic and local responses to the assault. The intensity of inflammation, coupled with local cellular events, also dictate therapeutic response and disease progression.

Although multiple factors influence the variable nature of organ involvement in lupus, in most situations autoantibodies participate in the initiation of disease activity. Nevertheless, debate continues over the properties of pathogenic antibodies, including how they form immune deposits and contribute to inflammation [2]. Early studies involving the Arthus reaction led to the notion that local immune complex formation within tissues was necessary for antibodies to initiate disease (reviewed in [3]). However, with the development of quantitative serum immune complex assays, general correlation of circulating levels with overall disease activity (initially in experimental rodent models and subsequently in human lupus) shifted the focus to deposition of circulating immune complexes as the proximate cause. It was postulated that the capacity of macrophages and other cells to remove complexes was either overwhelmed or impaired, and this led to complex deposition in tissues and inflammation [4]. Nevertheless, efforts to induce disease by passive administration of preformed immune complexes, of many shapes and sizes, to normal animals were unsuccessful, despite transient localization in various organs. Although these complexes sometimes activated inflammatory cellular programs in cultured cells, inflammation was not recapitulated in whole animals. Furthermore it was difficult to reconcile variable organ involvement among patients by this single mechanism.

Subsequently, it was discovered that immune deposits formed locally in serum sickness nephritis (the original poster child for deposition of circulating complex deposition) with antigen initially localizing in the kidney, followed by antibody binding, in situ [5]. The antigen's affinity for glomeruli was a major factor in the site of complex formation and subsequent inflammation. When more sophisticated methodologies became available, pathogenic autoantibodies were found to react directly with other tissue antigens in other experimental models of immune complex disease, suggesting that the antigen, whether endogenous or exogenous, determined both the site of deposition and the nature of organ involvement.

Application of these findings to human lupus was not immediate; however, evaluation of monoclonal anti-DNA antibodies, derived initially from lupus-prone mice, provided relevant insights. After transfer to normal animals, not all autoantibodies were pathogenic [6]. Furthermore, among the ‘pathogenic’ subset, individual antibodies were identified that had different pathological properties (e.g. they produced either nephritis, haemolytic anaemia, neurological disease or anti-phospholipid syndrome). Similar findings were made subsequently using human monoclonal autoantibodies [7]. These observations were consistent with clinical findings in patients with variable organ involvement, and they suggested that there might be subsets of human autoantibodies with different pathogenic properties. By extension, variable expression of pathogenic subsets among individuals could therefore contribute to differences in organ involvement.

An important clue to further understanding of the underlying mechanisms came from the observation that some anti-DNA antibodies cross-reacted with other autoantigens [8]. In some cases, such as with phospholipids, the reactivities were due to shared epitopes on these seemingly different molecules (e.g. the phosphodiester backbone shared by DNA and cardiolipin). In other situations the antigenic similarities were not readily apparent, and cross-reactivity was postulated to be due to either similar tertiary conformations on divergent molecules or/and a flexible antigen binding regions of the autoantibodies (i.e. induced fit). Although both mechanisms may be operative, the clinical implications of these findings were profound. They raised the possibility that lupus autoantibodies reacted directly with tissue antigens to form immune deposits. Furthermore, they implied that the site of deposition, or organ involvement, was determined by the presence of antibodies that reacted with either specific tissue antigens, or with endogenous antigens localized previously within tissues. In either scenario, the location of the tissue antigen dictated the site of deposition, and differences in autoantibody specificities (e.g. among patients) resulted in variation in organs involved. Identification of autoantibodies with specificity for tissue antigens only reinforced this viewpoint.

Many laboratories have since provided evidence to support an in situ mechanisms, with either antibodies binding to either organ-specific or circulating autoantigens that localize in tissues (reviewed in [2]). For example in the kidney, direct binding of autoantibodies to mesangial and glomerular endothelial cells, as well as matrix and basement membrane antigens, were demonstrated to initiate deposition and inflammation [9,10]. Additionally, the positively charged histone component of nucleosomes was observed to bind to negative charged moieties within the glomerular capillary wall and serve as a planted antigen for complex formation, with circulating anti-DNA and antinucleosome antibodies [11]. Thus, although the pro-inflammatory properties (e.g. isotype) of deposited autoantibodies influence the disease profile (e.g. through recruitment of inflammatory cells), it seems that the specific binding properties of autoantibodies dictate where the deposit forms initially, through direct interactions with endogenous antigens in tissues. Therefore, the antigen-binding properties of autoantibodies determine both which organs are involved, and where deposits form within them.

The findings of Hsieh et al. extend these observations to antineutrophil antibodies in human lupus, and they have important clinical implications [12]. The authors identified antibodies in lupus serum that reacted with SSB/La on the surface of neutrophils, and after binding to neutrophils some of the autoantibodies were internalized. Furthermore, of potential clinical relevance these autoantibodies inhibited phagocytosis, accelerated apoptosis and enhanced IL-8 production. Although the cellular mechanisms and pathways responsible for these events require elucidation, the findings support previous studies in murine lupus, and they suggest novel roles for these autoantibodies during the course of human disease. Particularly noteworthy, in addition to inciting inflammation, the antineutrophil antibodies effected neutrophil function directly through engagement of a specific neutrophil antigen. Whether cell surface events or antibody internalization were the principal mediators of the observed functional effects remains to be determined. Nevertheless, in addition to activation of inflammatory pathways through Fc mediated events (e.g. complement activation, cellular recruitment, etc.), these autoantibodies have the potential to render neutrophils less effective in combating infection by modulating their normal protective functions. It is easy to envision how similar interactions in patients could have dire effects and worsen the overall clinical situation. Similar examples of these types of antibody-mediated interactions in organ-specific autoimmune diseases, involving typically a single antigen, have been observed previously, and in some cases further evaluation of the consequences of autoantigen ligation per se has provided novel insights into either pathophysiology, normal function of the endogenous antigen or disease management [13–16]. In this regard, additional studies of anti SSB/La and other antibody–antigen interactions in lupus patients should clarify the pathological role of individual autoantibodies during disease, and the results have the potential to provide additional information relevant to physiological role of the autoantigens. As importantly, the studies should lead to the investigation of novel inflammatory pathways in lupus and other autoimmune diseases and the means to manipulate them.