C-reactive protein (CRP) is well known to rheumatologists. Levels of CRP in the blood serve as a reliable marker of disease activity in rheumatoid arthritis and various vasculitides. In systemic lupus erythematosus (SLE), the expression of CRP does not appear to correlate with disease activity, and levels generally remain low despite active disease (1). CRP is also known to infectious disease specialists; it serves as a marker of infection and participates in host defense (2). More recently, cardiologists have become intensely interested in CRP, and it has been widely examined as a risk factor for cardiovascular disease. These properties suggest that CRP is a sensitive marker for acute and chronic inflammation of diverse causes.
The biologic activities of CRP are less familiar to most clinicians but include the ability to activate the classical complement cascade (3), enhance phagocytosis (4), and bind to the Fcγ receptors (FcγR) (5, 6). In addition, the interaction of CRP with FcγR leads to cytokine production (7–9). The ligand-binding characteristics of CRP are also important in understanding its role in inflammation. In addition to recognition of microbial antigens, CRP reacts with cells at sites of tissue injury. CRP and the related molecule serum amyloid P component (SAP) bind to nuclear antigens, damaged membranes, and apoptotic cells (10, 11). Clearance of chromatin from the circulation is altered during the acute-phase response and is directly influenced by CRP and SAP (12). These studies have suggested that CRP and SAP are involved in the clearance of injured or apoptotic cells as well as the material released from these damaged cells.
The expression of CRP differs markedly between humans and mice. In humans, CRP is normally present at very low levels (usually less than 0.5 μg/ml) but typically increases to very high levels during the acute-phase response to infection or inflammation. Levels typically increase to 100–200 μg/ml and may increase to 1,000 μg/ml or higher in the case of severe tissue injury (e.g., burns). In the mouse, CRP is present at very low levels constitutively and increases to only 1–2 μg/ml during the acute-phase response. The transgenic mice used in the experiments conducted by Szalai et al, which are described elsewhere in this issue of Arthritis & Rheumatism (13), expressed human CRP as an acute-phase protein, although the levels of CRP in the blood remained low (∼1 μg/ml) throughout the life of the animal. The ability of the mice to produce high levels of CRP was verified by their strong increase in CRP levels when injected with lipopolysaccharide (LPS). Thus, the mouse model used resembles SLE in humans, in that patients with SLE have low levels of CRP unless they are exposed to an inflammatory signal, such as an infection.
Szalai and colleagues report that transgenic autoimmune mice expressing human CRP as an acute-phase serum protein are protected from nephritis and mortality (13). The human CRP gene, including the native human CRP gene promoter, was expressed in female (NZB × NZW)F1 mice, a model that is considered to most closely resemble SLE in humans. These findings are consistent with those of previous studies (14) and extend the observations by examining renal immune complex (IC) deposition and autoantibody profiles. The study by Szalai et al is important because it shows that CRP expressed at low serum levels prolongs the survival of mice with SLE. It is important to point out that CRP did not lower the levels of anti-DNA antibodies, and in fact increased the expression of IgG anti-DNA. The major protective effect of CRP in the study by Szalai et al was the marked decrease in IC deposition in the renal cortex but enhanced IC deposition in the mesangium. This finding suggests that CRP may not act to inhibit autoantibody production, but perhaps to regulate the clearance or processing of ICs. The investigators suggest a variety of possible mechanisms for the protective effect of CRP on autoimmunity.
Studies of the interaction of CRP with nuclear antigens began in the early 1980s, when it was determined that CRP bound to chromatin (15). It was later determined that CRP also bound to the D protein and the 70-kd protein of the small nuclear RNP (16, 17). (These proteins are the major targets of autoantibodies in patients with SLE and in patients with mixed connective tissue disease.) More recently, CRP has been shown to interact with apoptotic cells (10), although the interaction appears to be less avid than that of SAP, which is the major acute-phase pentraxin in the mouse. In a study by Bickerstaff et al (18), mice that were made genetically deficient in this protein developed autoimmune disease resembling SLE. Those investigators concluded that SAP acts in a manner similar to that of other molecules that bind to apoptotic cells, such as C1q, and that deficiency of SAP leads to a failure to properly clear nuclear antigens. SAP-deficient mice cleared chromatin from the circulation more rapidly than did normal mice. However, a direct relationship between SAP deficiency and nuclear antigen clearance in the pathogenesis of SLE remains unproven.
CRP or SAP could provide protection from autoimmune disease in several ways. The finding that CRP binds to nuclear antigens and apoptotic cells has led to speculation that CRP can enhance the clearance of nuclear antigens and prevent immunization by self antigens. It has, in fact, been demonstrated that CRP inhibits immunization by epitopes to which it binds (19). Since CRP is capable of interacting with FcγR, it is possible that CRP can alter the presentation of nuclear antigens by antigen-presenting cells. However, for CRP or SAP to provide protection against autoimmunity by this mechanism, it would presumably require the ongoing presence of CRP or SAP at elevated levels. In addition, the failure of CRP to lower autoantibody levels in the study by Szalai et al is not consistent with this hypothesis.
An alternative to the clearance hypothesis is suggested by several studies of the effects of CRP in models of inflammation. In those studies, initiated by Samols and colleagues (20, 21), transgenic mice expressing rabbit CRP were protected from pathologic inflammation in LPS-induced shock and alveolitis. The results of the experiments suggest that CRP induces antiinflammatory molecules that down-regulate the inflammatory cascade. Although CRP can induce the expression of tumor necrosis factor α and interleukin-1 (IL-1) (7), the ability of CRP to induce the production of the antiinflammatory cytokine IL-1 receptor antagonist may predominate (8). In the last several years, it has become apparent that CRP plays an active role in the regulation of inflammation and in the clearance of nuclear autoantigens. We have recently determined that the mechanism for the antiinflammatory activity of CRP in LPS-induced shock requires its interaction with FcγR (9). In this model, CRP was shown to induce the secretion of the antiinflammatory cytokine IL-10. This effect is analogous to the recently described ability of ICs to induce the release of the antiinflammatory cytokine IL-10 and to decrease the production of IL-12 (22, 23). In fact, a recent study by Szalai et al (24) showed increased levels of IL-10 in another experimental model of autoimmune disease.
Could induction of IL-10 contribute to the protective effect of CRP in SLE as well? Although IL-10 is a potent antiinflammatory cytokine, it also stimulates antibody production by B cells, and this may account for the elevated levels of anti-DNA antibodies seen in the study by Szalai et al (13). We have recently observed similar findings of increased antibodies to DNA in (NZB × NZW)F1 mice despite prolonged survival and decreased proteinuria in mice treated with exogenous CRP (25). Thus, it appears that CRP may act to decrease the inflammatory response to pathogenic ICs and anti-DNA antibodies. It is unclear whether this may be a local effect in the kidney, as suggested by Szalai and colleagues (13), or a more generalized effect on immune activation. However, the finding that CRP exerts a potent antiinflammatory effect in a wide variety of inflammatory states suggests that elevated levels of CRP produce a systemic down-regulation of the innate immune system.
Although CRP activates the complement cascade and many of its effects are dependent on complement activation, it is unclear whether CRP requires complement activation to mediate its protection from nephritis in SLE. CRP is a potent activator of the classical cascade of complement while inhibiting activation of the alternative pathway, as noted here. Szalai and colleagues suggest that complement activation by CRP may alter the site of deposition of ICs in the kidney (13). However, the central role of FcγR in the pathogenesis of lupus nephritis (26) would argue against the role of complement in CRP-mediated protection. The finding that CRP mediates protection from another inflammatory stimulus, endotoxin, through interaction with FcγR would favor an interaction between FcγR and CRP in the regulation of lupus nephritis. Studies examining the effect of CRP on nephritis with respect to complement and FcγR deficiency will help to clarify this point.
Szalai and colleagues' study, despite its intriguing and important findings, is clearly the beginning of the story. We are left with as many questions as answers. The ultimate role of CRP and SAP in the regulation of a variety of inflammatory conditions will require substantial new studies to further clarify the activities of CRP that are important for its biologic effects. Because of the ability of CRP to clear nuclear antigens, activate complement, and promote phagocytosis and cytokine production through interactions with FcγR, dissecting the role of these activities under a variety of conditions (e.g., systemic elevation of CRP levels versus localized CRP deposition and variable levels of FcγR expression on phagocytic cells) will require further investigation.
These studies suggest that the pentraxins CRP and SAP, as well as other acute-phase proteins, act to regulate the immune system, perhaps in many ways. The interaction of CRP with a variety of mediators and receptors will likely lead to both pro- and antiinflammatory activities under different conditions of cell activation, tissue deposition, and ligand availability. The exploitation of these natural regulators of inflammation and autoimmunity may eventually provide a more physiologic and less toxic approach to the treatment of a broad range of medical conditions, including infection, inflammation, and cardiovascular disease.