Uncoupling of Immune Complex Formation and Kidney Damage in Autoimmune Glomerulonephritis

See allHide authors and affiliations

Science  13 Feb 1998:
Vol. 279, Issue 5353, pp. 1052-1054
DOI: 10.1126/science.279.5353.1052


The generation of autoantibody and subsequent tissue deposition of immune complexes (IC) is thought to trigger the pathogenic consequences of systemic autoimmune disease. Modulation of the autoantibody response disrupts pathogenesis by preventing the formation of ICs; however, uncoupling IC formation from subsequent inflammatory responses seems unlikely because of the apparent complexity of the IC-triggered inflammatory cascade. However, the disruption of a single gene, which encodes the γ chain of the Fc receptor, was found to achieve this uncoupling in a spontaneous model of lupus nephritis, the New Zealand Black/New Zealand White (NZB/NZW) mouse. Gamma chain–deficient NZB/NZW mice generated and deposited IC and activated complement, but were protected from severe nephritis, thus defining another potential pathway for therapeutic intervention in autoimmune disease.

NZB mice develop autoantibodies and autoimmune hemolytic anemia, but show no signs of glomerular disease until crossed to the NZW background to generate NZB/NZW (B/W F1) mice (1). A minimum of three distinct genetic loci are required for the manifestation of autoimmune glomerulonephritis in the B/W F1, two derived from NZB and one from NZW mice (2, 3). Several features of this model are consistent with lupus in humans. Females develop disease at a frequency 10 times that of males, and IC and complement deposition in glomeruli are observed. Significant proteinuria is seen concomitant with the serological appearance of antibodies to DNA as well as ICs of the immunoglobulin G1 (IgG1), IgG2a, and IgG2b subclasses beginning at 4 months (1). Median survival is 6 months, with mortality resulting from renal failure. Several studies have demonstrated the essential role of B cells (4) and autoantibodies (5, 6) in disease development. Agents that interfere with autoantibody production have been shown to attenuate disease (7-12). Disruption of the subsequent inflammatory response triggered by glomerular IC deposition represents an alternative therapeutic approach, but success may be complicated by the large number of possible proinflammatory molecules presumed to be activated by ICs, including complement components and the cellular receptors for IgG. Complement depletion attenuates disease in several induced models of glomerulonephritis (13, 14), and anti-C5a treatment modulates glomerular injury in a spontaneous murine model (15). However, the primacy of complement activation in IC-triggered inflammation has been questioned by several recent genetic studies (16-20).

In the classical model of IC-triggered inflammation, the Arthus reaction, the demonstration that FcγRs are essential whereas complement is not (16-20) suggests that the fundamental assumption of the pathogenesis of autoimmune glomerulonephritis as being mediated by complement activation (13, 2127) requires reevaluation. Such studies have been facilitated by the availability of defined murine strains deficient in components of this pathway. Mouse strain γ / , which is deficient in the Fc receptor (FcR) γ chain, does not express the activation receptors FcγRI and FcγRIII, but still bears the inhibitory receptor FcγRIIB. To determine if the spontaneous autoimmune glomerulonephritis in the B/W F1 required FcγRs for disease development, NZB and NZW mice were backcrossed for eight generations to the γ / mouse strain, and animals homozygous or heterozygous for the disruption in the γ chain were identified. We obtained γ+/ and γ / B/W F1animals and observed them for evidence of autoimmune glomerulonephritis. As reported (1), B/W F1 animals have a median survival of 200 days, succumbing to the sequelae of renal failure (Fig.1A). In contrast, γ / B/W mice had a prolonged survival, with 82% (20 of 22) alive at 9 months. In B/W mice, the appearance of proteinuria presages the onset of clinically severe disease, with mortality following within weeks. In contrast, γ / B/W animals have a delayed onset and a reduced incidence of proteinuria (P< 0.00012) compared to their heterozygous littermates (Fig. 1B). In addition, the onset of proteinuria in the γ / B/W animals did not correlate with mortality; γ / B/W animals with significant proteinuria did not progress to the sequelae of glomerulonephritis and renal failure (28).

Figure 1

(A) γ / enhances B/W survival. Kaplan-Meier cumulative proportion survival of γ+/ and γ / B/W F1female mice (45). A total of 31 γ / and 39 γ+/ B/W F1 mice were followed. Median survival of γ+/ B/W mice was 200 ± 7 days; median survival of γ / B/W mice was 400 days. Premorbid γ+/ B/W animals were uremic, anemic, cachectic, and edematous. (B) Delayed onset and diminished severity of proteinuria in γ / B/W mice. Severe proteinuria (>5 mg/ml) (46) developed in all γ+/ B/W mice, with a median age of onset of 188 ± 11 days. In contrast, the disease incidence, severity, and time of onset were attenuated in γ / B/W mice. Normalized data are expressed as: total proteinuria = (incidence) × (mean 24-hour urinary protein).

To determine the molecular basis for protection, γ / and γ+/ B/W mice were assayed for the presence of antibodies to double-stranded DNA (dsDNA) and circulating ICs (Table 1). No significant differences were seen in isotype, specificity, or total immunoglobulin between these animals, indicating that the backcrossing had transferred the B/W loci responsible for autoantibody production. In addition, lack of FcγRI and FcγRIII did not affect the clearance of soluble ICs, with similar steady-state levels of circulating complexes seen in both genotypes. In contrast, clearance of insoluble ICs—such as antibody-coated red blood cells, platelets, or microbial pathogens—is FcR-dependent (20, 29). Production of ICs in γ / B/W mice were not expected to differ, because B and T cell responses are normal in γ / mice (30,32).

Table 1

Titers of ICs and antibodies to dsDNA for γ+/– and γ–/– mice. Results are expressed as a percent of a pooled serum from 7-month-old B/W female mice used to create a standard curve (48). Means ± SEM of two separate experiments of 21 heterozygous and 29 homozygous deficient female animals 6 to 7 months of age are shown.

View this table:

Glomerular samples taken from γ−/− and γ+/− B/W mice were examined histologically and by immunofluorescence. Active glomerular disease was seen in the γ+/ B/W animals, including mesangial thickening and hypercellularity evolving into end-stage crescent formation and sclerosis (Fig. 2). IC deposition was observed in both genotypes, along with deposition of complement C3. However, age-matched γ / B/W showed no evidence of inflammatory disease; only mesangial thickening was seen, indicative of IC deposition revealed by immunofluorescence (Fig.3). Such deposition is seen only in the B/W background, because γ / animals on a variety of backgrounds do not generate or deposit IC or C3 in their kidneys (33). Thus, despite the deposition of ICs and C3 in the glomeruli of γ / B/W, the inflammatory response was uncoupled, indicating that the presence of FcγRs was required, and complement activation by ICs is not sufficient to initiate an inflammatory cascade.

Figure 2

Glomerulonephritis is blocked in γ / B/W mice. Representative renal glomeruli (47) of C57Bl/6 γ / , B/W γ / , and B/W γ+/ 7-month-old female mice (magnification ×400). Pathological features of B/W γ+/ glomeruli include mesangial thickening and hypercellularity evolving into end-stage sclerotic and crescentic changes. C57Bl/6 γ / glomeruli demonstrate normal glomerular architecture, and B/W γ / glomeruli show relatively mild mesangial thickening without concomitant inflammatory changes. Despite the development of mesangial thickening in B/W γ / mice, there is little hypercellularity and no evidence of end-stage glomerular changes.

Figure 3

IC and complement deposition are similar in γ+/ and γ / B/W mice. Immunohistochemical analysis (47) of kidneys from the animals shown in Fig. 2. Both mice had comparable concentrations of circulating ICs and autoantibodies to dsDNA; however, proteinuria was evident in the γ+/ , but not the γ / B/W mouse. Prominent mesangial IgG and C3 staining is evident in B/W mice of both genotypes.

An induced model of IC-mediated glomerulonephritis done in γ / -deficient mice (34) confirmed our results. Rabbit antibodies to glomerular basement membrane, when passively transferred into wild-type mice sensitized to rabbit IgG, developed an acute glomerulonephritis. In the absence of γ chain, reduction in disease was seen. Complement depletion with cobra venom factor did not attenuate disease, supporting the data shown in Fig. 3.

The likely receptor involved in IC triggering is FcγRIII, whose expression on mesangial cells may be essential for disease initiation (35, 36). The similarities in serum IC levels and deposition in γ / and wild-type mice suggest that these receptors have a minimal contribution to the clearance of ICs, indicating that IC clearance is either mediated by FcγRII, complement receptors, or both. These divergent roles for FcRs and complement in autoimmune disease offer an explanation for the apparent paradox that deficiencies in complement increase the risk of lupus (37, 38). The complement system may regulate autoreactive B cells (39) as well as contribute to IC clearance, whereas FcγRIII mediates the inflammatory activation by ICs. Thus, deficiencies in complement would result in an increase in autoantibodies and a reduction of IC clearance with a corresponding increase in IC deposition, thereby increasing FcR-mediated activation. These studies and ours indicate that complement and Fc receptors have evolved for distinctly different roles in their interaction with ICs. Complement has been shown to be essential for innate immunity against microbial pathogens, requiring natural antibodies to mediate their protective effect (40, 42), whereas FcγRs have emerged as the principal system for coupling antigen-specific IgG antibodies to cellular effector responses and play a minor role in host innate immunity (29, 43, 44). Therefore, this distinction argues for the development of new therapeutic strategies based on FcR blockade for the treatment of autoimmune glomerulonephritis.

  • * To whom correspondence should be addressed. E-mail: ravetch{at}


View Abstract

Navigate This Article