Many forms of glomerulonephritis are triggered by Ab localization in the glomerulus, but the mechanisms by which this induces glomerular inflammation are not fully understood. In this study we investigated the role of complement in a mouse model of cryoglobulin-induced immune complex glomerulonephritis. Several complement-deficient mice on a C57BL/6 and BALB/c genetic background were used and compared with strain-matched, wild-type controls. Cryoglobulinemia was induced by i.p. injection of 6-19 hybridoma cells producing an IgG3 cryoglobulin with rheumatoid factor activity against IgG2a of allotype a present in BALB/c, but not C57BL/6, mice. Thus, the cryoprecipitate in C57BL/6 mice consisted of the IgG3 cryoglobulin only (type I cryoglobulinemia) compared with IgG3-IgG2a complexes in BALB/c (type II cryoglobulinemia). The survival of mice was not affected by complement deficiency. Glomerular influx of neutrophils was significantly less in C3-, factor B-, and C5-deficient mice compared with wild-type and C1q-deficient mice. It did not correlate with C3 deposition, but did correlate with the amount of C6 deposited. Deficiency of CD59a, the membrane inhibitor of the membrane attack complex, did not induce an increase in neutrophil infiltration, suggesting that the generation of C5a accounts for the effects observed. There was no apparent difference between cryoglobulinemia types I and II regarding the role of complement. Our results suggest that in this model of cryoglobulin-induced glomerulonephritis the neutrophil influx was mediated by C5 activation with the alternative pathway playing a prominent role in its cleavage. Thus, blocking C5 is a potential therapeutic strategy for preventing renal injury in cryoglobulinemia.

Immune complex glomerulonephritis is an important feature of a number of human diseases, including cryoglobulinemia, systemic lupus erythematosus, serum sickness, and bacterial endocarditis. It is characterized by the formation and/or deposition of Ag-Ab complexes in the glomerulus, followed by an influx of inflammatory leukocytes. The mechanisms inducing the glomerular inflammation and the subsequent tubulointerstitial injury are not well understood.

Complement is well known to mediate the processing of immune complexes (1). However, the relative importance of complement activation by immune complexes in the induction of immune complex-mediated disease manifestations remains unclear. Although the role of the complement system has been investigated in a number of experimental models of immune complex glomerulonephritis using various strategies, the findings are still conflicting. Since the early studies performed by Unanue and Dixon (2), complement has been considered to mainly play a proinflammatory role in the glomerulus. However, recent studies have shown that mice lacking C1q, the first component of the classical pathway of complement activation, develop a more severe glomerular inflammation and thrombosis than their wild-type controls in the accelerated nephrotoxic nephritis model, suggesting that the role of complement may be protective (3).

Cryoglobulinemia is caused by Igs that precipitate in the cold. Depending on the clonality of the precipitating Igs, three types of cryoglobulinemia can be distinguished (4). Although type I cryoglobulinemia consists of a single monoclonal cryoprecipitable Ig, type II and III cryoglobulins are mixed, i.e., composed of monoclonal (type II) or polyclonal (type III) Abs with rheumatoid factor activity against other Igs. In 1987, Gyotoku et al. (5) described a mouse model of cryoglobulinemia using a cryoprecipitating IgG3 6-19 mAb derived from autoimmune MRL/MpJ-lpr/lpr mice. This Ab had additional rheumatoid factor activity against IgG2a of allotype a. Therefore, depending on the IgG2a allotype, injected mice developed type I or type II cryoglobulinemia. Both types of cryoglobulinemia led to severe glomerular injury with predominant infiltration of polymorphonuclear neutrophils (6), as may sometimes be seen in human disease (7). Additional studies of this model of cryoglobulinemia demonstrated that the glomerular inflammation was dependent on cryoprecipitation of the monoclonal component, but not on its rheumatoid factor activity (8). The infiltrating polymorphonuclear neutrophils played an active role in the development of the wire-loop glomerular lesions observed in this model (9), and the inflammation was not mediated by FcγRs (10) or C3 (11).

In the present study we analyzed the role of complement in the pathogenesis of the initial glomerular inflammation induced by 6-19 IgG3 monoclonal cryoglobulins. We used two different strains of mice developing either type I (C57BL/6 mice) or type II (BALB/c mice) cryoglobulinemia. Our results suggest that C5 had a predominant role in neutrophil recruitment. Furthermore, C5 cleavage occurred only in mice with a fully functional alternative pathway. Interestingly, the glomerular inflammation was independent of glomerular C3 deposition, because C3 deposition occurred via the classical and the alternative pathway to a similar extent.

Complement-deficient (C1qa, C3, Bf, CD59a, Bf/C2, C4) mice were generated as described previously (12, 13, 14). C5-deficient mice were generated by backcrossing the mutated C5 gene present in DBA/2 mice into C57BL/6 mice for 10 generations. Age-, strain-, and sex-matched, wild-type mice were used in all experiments. Complement-deficient mice studied on the C57BL/6 genetic background were backcrossed to that strain for 10 generations. On the BALB/c genetic background, only C1q-, C3-, and CD59a-deficient mice were available. Mice lacking other complement components, specifically factor B, C2, and C4, were not investigated on the BALB/c genetic background, because these complement genes are part of the H2 region. Therefore, even after extensive backcrossing there would be an MHC mismatch with the backcrossed mice carrying the H2b haplotype of the original 129 embryonic stem cells instead of the H2d haplotype of wild-type BALB/c controls. All animal procedures were performed in accordance with institutional guidelines.

Cryoglobulinemia was induced using a protocol modified from that described previously (5). Briefly, 107 hybridoma cells producing a cryoprecipitating murine IgG3 Ab 6-19 were injected i.p. without pretreatment with pristane. To avoid rejection of the hybridoma cells, immunosuppression was achieved by a simultaneous injection of a mixture of anti-mouse CD4 (GK 1.5) and anti-mouse CD8 (H-35) mAbs. In concordance with previous reports (10, 15, 16, 17), disease manifestations occurred within 1–2 wk after the injection of hybridoma cells. For more detailed analyses, mice were killed at various time points before or at the onset of severe signs of disease.

For the purification of cryoglobulins, blood without addition of anticoagulant was kept at 37°C until a clot had established and serum could be separated. Serum (100 μl) was incubated in glass tubes for 4–7 days at 4°C. Tubes were then centrifuged, and the pellets were washed twice with PBS plus 0.05% Tween 20. Cryoprecipitates were eventually resuspended in 100 μl of 4 M urea. Resolubilized cryoglobulins were quantified by ELISA (see below). IgG3 cryoglobulin concentrations in serum were also measured.

For ELISA, 96-well microtiter plates (Nunc-Immuno MaxiSorp; Nunc) were coated with 50 μl of polyclonal goat anti-mouse Ig H and L chain Ab (Southern Biotechnology Associates) diluted 1/1000 in sodium carbonate buffer (pH 9.6), and incubated at 4°C overnight. Plates were blocked with 100 μl of PBS/0.5% BSA. Sera or resolubilized cryoglobulins were appropriately diluted in PBS/2.0% BSA/0.05% Tween 20 and incubated for 1 h at 37°C in duplicate. Each assay included affinity-purified mouse IgG3 (Sigma-Aldrich), titrated to generate standard curves. Plates were washed and further incubated for 1 h at 37°C with alkaline phosphatase-conjugated, goat anti-mouse IgG3-specific Ab (Southern Biotechnology Associates) diluted 1/1000 in PBS/2.0% BSA/0.05% Tween 20. The plates were developed using p-nitrophenyl phosphate (Sigma-Aldrich) as substrate. The OD of the reaction mixture at the 405-nm wavelength was measured using an ELISA reader (Titer-Tek Labsystems). The relative concentrations of IgG3 in individual samples were calculated by comparing the mean OD obtained for duplicate wells minus nonspecific binding to the titrated mouse IgG3 standard curve.

Serum creatinine was measured using an Olympus AU600 autoanalyzer (Olympus Diagnostics). To quantify albuminuria, mice were housed in metabolic cages for 24 h for urine collection. The albumin concentration was measured by radial immunodiffusion. Samples and standards (mouse albumin; Sigma-Aldrich) were placed in wells (4 μl/well) in 1.2% agarose in PBS containing rabbit anti-mouse albumin (Biogenesis). Gels were dried and stained with Coomassie Blue. The albumin concentration was calculated with reference to a standard curve.

Kidneys were fixed for 2 h in Bouin’s solution, transferred to 70% ethanol, and embedded in paraffin wax. Sections were cut and then stained with periodic acid-Schiff. The numbers of neutrophils per glomerulus were expressed as the mean count for 25 glomeruli. All analyses were performed blind to sample identity.

Kidneys were snap-frozen in isopentane and stored at −70°C. Frozen sections were cut at a thickness of 5 μm. An observer blind to the sample identity performed all quantitative immunofluorescence analyses. Rabbit anti-mouse IgG (FITC-conjugated; Southern Biotechnology Associates), goat anti-mouse IgG3 (FITC-conjugated; Serotec), and goat anti-mouse C3 (FITC-conjugated; ICN) Abs were used for direct immunofluorescence studies. All incubations were performed for 1 h at room temperature, and all Abs were appropriately diluted in PBS. Sections were mounted in Permafluor. In quantitative immunofluorescence studies, to exclude artifacts due to variable decay of the fluorochrome, all sections from one experiment were stained and analyzed at the same time. Sections were examined at ×40 magnification using a BX4 fluorescence microscope (Olympus Optical). A Color Coolview digital camera (Photonic Science) was attached to the microscope, and using Image-Pro Plus software (Media Cybernetics), images were captured for analysis. For each section, 20 glomeruli were examined, and the mean fluorescence intensity was recorded, with results expressed as arbitrary fluorescence units (AFI).4

C6 staining was conducted by incubation of the sections with a rabbit anti-mouse C6 Ab (18) (provided by Dr. A. Tenner, Department of Molecular Biology and Biochemistry, University of California, Irvine, CA) diluted 1/400 in TBS/0.1% BSA for 60 min. Ag-bound Abs were detected using a secondary goat anti-rabbit IgG HRP-labeled Ab (DakoCytomation). The sections were developed with 3,3′-diaminobenzidine (Sigma-Aldrich), counterstained with Mayer’s hematoxylin solution (Sigma-Aldrich), and then dehydrated through graded alcohols and xylenes. C6-deficient mice (provided by Dr. P. Morgan, Department of Biochemistry, University of Wales College of Medicine, Cardiff, U.K.) were used as negative controls. For a quantitative assessment of glomerular C6 deposition, the observer was blinded to sample identity, and the intensity of glomerular staining was ranked with 0 indicating the lowest staining intensity.

For the quantification of glomerular macrophages, kidneys were fixed in paraformaldehyde-lysine-periodate for 4 h, then transferred to 7% sucrose overnight before snap-freezing in isopentane and storage at −70°C. For the staining, a primary monoclonal rat anti-mouse CD68 Ab (Serotec) was used. Sections were blocked with a 1% solution of hydrogen peroxide in 50% methanol. A mouse anti-rat secondary Ab and a rat peroxidase anti-peroxidase tertiary Ab (both purchased from Jackson ImmunoResearch Laboratories) were then applied. The sections were developed with 3,3′-diaminobenzidine, counterstained with Mayer’s hematoxylin solution, and then dehydrated through graded alcohols.

All values described in the text and figures are expressed as the median and range. Statistical analysis was conducted using PRISM 3.2 (GraphPad). Nonparametric tests were applied throughout, with differences considered significant at p < 0.05.

The i.p. implantation of hybridoma cells secreting the cryoprecipitating murine IgG3 6-19 induced proliferative glomerulonephritis with prominent endocapillary hypercellularity and neutrophil infiltration (Fig. 1 A). This was followed by systemic disease manifestations. The mice usually developed severe signs of disease between 8 and 12 days after injection. As expected, complement-deficient mice had no survival advantage over wild-type controls (data not shown). Renal disease was usually best seen on days 7–8 after the hybridoma injection and before the development of other clinical signs. Therefore, this time point was chosen for all additional analyses. Mice that rejected the hybridoma cells were excluded from further analysis.

FIGURE 1.

Light microscopy of glomeruli 7 days after peritoneal implantation of hybridoma cells in mice on a C57BL/6 background. Wild-type mice (A) showed proliferative glomerulonephritis with enlargement of the glomerular tuft and prominent infiltration of neutrophils. C1q-deficient mice showed a similar appearance (B). In mice lacking factor B (C) or C3 (D), there was less glomerular hypercellularity and fewer neutrophils in the glomeruli (periodic acid-Schiff stain; magnification, ×40).

FIGURE 1.

Light microscopy of glomeruli 7 days after peritoneal implantation of hybridoma cells in mice on a C57BL/6 background. Wild-type mice (A) showed proliferative glomerulonephritis with enlargement of the glomerular tuft and prominent infiltration of neutrophils. C1q-deficient mice showed a similar appearance (B). In mice lacking factor B (C) or C3 (D), there was less glomerular hypercellularity and fewer neutrophils in the glomeruli (periodic acid-Schiff stain; magnification, ×40).

Close modal

The model of cryoglobulinemia was first analyzed in mice with a C57BL/6 genetic background. These mice developed type I cryoglobulinemia. As shown in Figs. 1 and 2,A, C3- and factor B-deficient mice had significantly lower glomerular neutrophil influx compared with wild-type mice, whereas no difference was seen in C1q-deficient mice. In parallel with the reduced numbers of glomerular neutrophils, C3- and factor B-deficient mice also had lower serum creatinine concentrations, although the difference from wild-type mice was only significant for factor B-deficient mice (Fig. 2,B). Interestingly, in neither of the groups did albuminuria or hematuria differ significantly from the prediseased stage (data not shown). The differences seen in neutrophil influx and serum creatinine were not due to variation in IgG3 serum concentrations, cryoprecipitating IgG3, or reduced glomerular deposition of the IgG3 cryoglobulin (Table I). In addition, no significant differences in glomerular macrophage number between wild-type and factor B-deficient mice were detected. The median number of macrophages (range) was 4.9 (2.1–10.6) in wild-type and 5.5 (3.4–6.5) in factor B-deficient mice. Therefore, the results suggested a prominent role for the alternative pathway of complement in the pathogenesis of cryoglobulin-induced glomerular neutrophil infiltration.

FIGURE 2.

Type I cryoglobulinemia in wild-type and C1q-, C3-, and factor B-deficient C57BL/6 mice 7 days after the injection of 6-19 hybridoma cells. A, Mean number of neutrophils per glomerulus. Although C1q-deficient mice showed similar neutrophil influx compared with wild-type controls, the number of glomerular neutrophils was significantly reduced in C3- and factor B-deficient mice. B, Serum creatinine. Serum creatinine concentrations in factor B-deficient mice were significantly lower than in controls. The slightly lower concentrations seen in C3-deficient mice did not differ significantly from those in wild-type mice. Data shown are from a representative experiment conducted at least twice. The horizontal bar represents the median of each group.

FIGURE 2.

Type I cryoglobulinemia in wild-type and C1q-, C3-, and factor B-deficient C57BL/6 mice 7 days after the injection of 6-19 hybridoma cells. A, Mean number of neutrophils per glomerulus. Although C1q-deficient mice showed similar neutrophil influx compared with wild-type controls, the number of glomerular neutrophils was significantly reduced in C3- and factor B-deficient mice. B, Serum creatinine. Serum creatinine concentrations in factor B-deficient mice were significantly lower than in controls. The slightly lower concentrations seen in C3-deficient mice did not differ significantly from those in wild-type mice. Data shown are from a representative experiment conducted at least twice. The horizontal bar represents the median of each group.

Close modal
Table I.

Total serum IgG3 concentrations, cryoprecipitating IgG3 and glomerular IgG3 deposition on day 7 after injectiona

StrainC57BL/6 Genetic Background
Wild typeC1qa−/−C3−/−Bf−/−
Total serum IgG3 (mg/ml) 1.26 (0.64–1.59) 1.0 (0.84–1.84) 0.84 (0.52–1.84) 0.91 (0.68–1.89) 
Cryoprecipitating IgG3 (mg/ml) 0.59 (0.2–5.8) 0.48 (0.4–1.1) 0.69 (0.3–2.9) 1.75 (0.7–2.9) 
Glomerular IgG3 deposition (AFI) 27.8 (23.7–32.0) 31.5 (23.5–46.1) 31.6 (22.3–43.6) 31.3 (23.0–37.1) 
StrainC57BL/6 Genetic Background
Wild typeC1qa−/−C3−/−Bf−/−
Total serum IgG3 (mg/ml) 1.26 (0.64–1.59) 1.0 (0.84–1.84) 0.84 (0.52–1.84) 0.91 (0.68–1.89) 
Cryoprecipitating IgG3 (mg/ml) 0.59 (0.2–5.8) 0.48 (0.4–1.1) 0.69 (0.3–2.9) 1.75 (0.7–2.9) 
Glomerular IgG3 deposition (AFI) 27.8 (23.7–32.0) 31.5 (23.5–46.1) 31.6 (22.3–43.6) 31.3 (23.0–37.1) 
a

Data are expressed as the median, with the range of values in parentheses. There was no significant difference between the groups of mice.

To analyze whether the reduced neutrophil influx in alternative pathway-deficient mice (i.e., factor B- and C3-deficient mice) was due to reduced complement deposition, kidney sections were stained for C3 and C6. Although there was a trend toward less glomerular C3 deposition in C1q- and factor B-deficient mice, there was no overall difference compared with wild-type mice (Fig. 3,A). Furthermore, C3 deposition in C1q- and factor B-deficient mice was very similar, suggesting that both pathways of complement activation were activated by the deposited cryoglobulins. However, in contrast to the C3 deposition, staining for glomerular C6 reflected the differences in glomerular neutrophil influx, with significantly less C6 deposition in the C3- and factor B-deficient mice compared with wild-type controls (Fig. 3 B). The presence of very low level of C6 deposits in C3-deficient mice may represent local C6 synthesis.

FIGURE 3.

Glomerular C3 and C6 deposition in wild-type and C1q-, C3-, and factor B-deficient C57BL/6 mice 7 days after injection. A, The AFI of the C3 deposition in C1q- and factor B-deficient mice was slightly, but not significantly, lower than that in controls. C3-deficient mice had only background staining. Staining below the dashed line was considered background staining. B, Ranking of glomerular C6 deposition by an observer blinded to sample identity revealed a similar staining intensity between C1q-deficient and wild-type mice. However, factor B-deficient mice had significantly less C6 deposition compared with controls. The very low level of C6 staining found in C3-deficient mice most likely represented local C6 synthesis. The horizontal bar represents the median of each group.

FIGURE 3.

Glomerular C3 and C6 deposition in wild-type and C1q-, C3-, and factor B-deficient C57BL/6 mice 7 days after injection. A, The AFI of the C3 deposition in C1q- and factor B-deficient mice was slightly, but not significantly, lower than that in controls. C3-deficient mice had only background staining. Staining below the dashed line was considered background staining. B, Ranking of glomerular C6 deposition by an observer blinded to sample identity revealed a similar staining intensity between C1q-deficient and wild-type mice. However, factor B-deficient mice had significantly less C6 deposition compared with controls. The very low level of C6 staining found in C3-deficient mice most likely represented local C6 synthesis. The horizontal bar represents the median of each group.

Close modal

Taken together, these results indicated that there was no direct pathogenic role for C3 and that neutrophil recruitment was dependent on activation of the late complement components predominantly via the alternative pathway. To provide additional support for this observation, glomerular neutrophil infiltration in diseased mice was compared between wild-type and C5-deficient mice. As seen for C3- and factor B-deficient mice, C5 deficiency led to significantly reduced glomerular neutrophil infiltration, confirming the suggested role for the late complement components in neutrophil recruitment (Fig. 4). This could be mediated by the membrane attack complex (C5b-9) or by the chemotactic properties of cleaved C5. To address this question, we analyzed mice deficient in CD59a, the cell surface inhibitor of the membrane attack complex. Compared with wild-type controls, similar numbers of glomerular neutrophils could be observed (wild-type mice: median number of neutrophils, 2.2; range, 0.8–7.2; CD59a-deficient mice: median number of neutrophils, 1.8; range, 0.3–5.6; p = 0.7), supporting a major role for C5a in neutrophil recruitment in this model of cryoglobulin-induced immune complex glomerulonephritis.

FIGURE 4.

Neutrophil recruitment in wild-type and C5-deficient C57BL/6 mice developing type I cryoglobulinemia 7 days after injection of hybridoma cells. The number of neutrophils was significantly reduced in C5-deficient mice. The horizontal bar represents the median of each group.

FIGURE 4.

Neutrophil recruitment in wild-type and C5-deficient C57BL/6 mice developing type I cryoglobulinemia 7 days after injection of hybridoma cells. The number of neutrophils was significantly reduced in C5-deficient mice. The horizontal bar represents the median of each group.

Close modal

In a second step, the model of cryoglobulinemia was applied to mice on a BALB/c genetic background. Because BALB/c mice have an IgG2a of allotype a, the cryoprecipitating IgG3 had additional rheumatoid factor activity leading to a coprecipitation of monoclonal IgG3 and polyclonal IgG2a. Therefore, in these mice the model resembled type II cryoglobulinemia.

As seen in C57BL/6 mice, C3-deficient mice on a BALB/c genetic background were also protected from severe glomerular neutrophil influx, whereas C1q-deficient mice were similar to the wild-type controls (Fig. 5,A), indicating that the alternative pathway played a predominant role even in this model of disease. The differences observed could not be attributed to variation in glomerular IgG deposition, because this parameter did not significantly differ among the three groups (Table II). Furthermore, the AFI for glomerular C3 deposits did not significantly differ between C1q-deficient and wild-type mice (Table II). However, as in type I cryoglobulinemia, C6 deposition strongly reflected the number of infiltrated neutrophils, with almost undetectable levels in C3-deficient mice (Fig. 5 B). Therefore, although it was not possible to test factor B-deficient mice on a BALB/c genetic background, the same prominent role of the alternative pathway of complement as that seen in mice with type I cryoglobulinemia (C57BL/6) was likely. In addition, we were able to address the question of whether neutrophil influx is due to an effect of the membrane attack complex (C5b-9) or to the chemotactic properties of cleaved C5 by analyzing mice deficient in CD59a. Similar to the results obtained in C57BL/6 mice, the comparison of CD59a-deficient and wild-type mice did not reveal a significant difference in the number of glomerular neutrophils. The median number of neutrophils was 6.2 (range, 0.5–8.2) in wild-type mice and 5.2 (range, 3.2–7.1) in CD59a-deficient mice, again supporting a major role for C5a in neutrophil recruitment in this model of cryoglobulin-induced immune complex glomerulonephritis.

FIGURE 5.

Type II cryoglobulinemia in wild-type and C1q- and C3-deficient BALB/c mice 8 days after injection of 6-19 hybridoma cells. A, Mean number of neutrophils per glomerulus. Although C1q-deficient mice showed similar neutrophil influx compared with wild-type controls, the glomerular neutrophils were significantly reduced in C3-deficient mice. B, Ranking of glomerular C6 deposition revealed a similar staining intensity between C1q-deficient and wild-type mice. C3-deficient mice had almost undetectable C6 deposition compared with controls. The horizontal bar represents the median of each group.

FIGURE 5.

Type II cryoglobulinemia in wild-type and C1q- and C3-deficient BALB/c mice 8 days after injection of 6-19 hybridoma cells. A, Mean number of neutrophils per glomerulus. Although C1q-deficient mice showed similar neutrophil influx compared with wild-type controls, the glomerular neutrophils were significantly reduced in C3-deficient mice. B, Ranking of glomerular C6 deposition revealed a similar staining intensity between C1q-deficient and wild-type mice. C3-deficient mice had almost undetectable C6 deposition compared with controls. The horizontal bar represents the median of each group.

Close modal
Table II.

Serum creatinine concentrations and AFI of glomerular IgG and C3 deposition 7 days after the injection of the hybridoma cellsa

StrainBALB/c Genetic Background
Wild typeC1qa−/−C3−/−
Serum creatinine (μmol/L) 44 (37–47) 43 (39–52) 36 (31–46) 
Glomerular IgG deposition (AFI) 25.4 (14.7–33.9) 20.7 (16.0–26.5) 29.5 (22.6–40.2) 
Glomerular C3 staining (AFI) 64.6 (36.4–92.9) 77.1 (42.8–97.9) N/A 
StrainBALB/c Genetic Background
Wild typeC1qa−/−C3−/−
Serum creatinine (μmol/L) 44 (37–47) 43 (39–52) 36 (31–46) 
Glomerular IgG deposition (AFI) 25.4 (14.7–33.9) 20.7 (16.0–26.5) 29.5 (22.6–40.2) 
Glomerular C3 staining (AFI) 64.6 (36.4–92.9) 77.1 (42.8–97.9) N/A 
a

Data are expressed as the median, with the range of values in parentheses. N/A, not applicable.

In this study we demonstrated that complement plays a predominant role in neutrophil recruitment in a model of cryoglobulin-induced immune complex glomerulonephritis. Neutrophils have been shown to be the major inflammatory cells, which are initially observed at the sites of tissue injury induced by 6-19 monoclonal cryoglobulins including the kidney (6). Although neutrophils played an active role in the development of wire-loop glomerular lesions (9), how the glomerular deposits of 6-19 cryoglobulins led to neutrophil infiltration remained unclear (11). In the previous study it was shown, using cobra venom factor to deplete murine C3, that complement deposition did not play a major role in the development of the wire-loop lesions, but glomerular neutrophil infiltration was not quantitated in the C3-depleted animals. Our study using complement-deficient mice, demonstrated that there was an important role for complement in the recruitment of neutrophils to the glomeruli. There are, however, some differences between the studies. In our study we did not pretreat the animals with pristane, because this might cause an additional and independent effect on complement activation. Consequently the cryoglobulin concentrations achieved in our model were markedly lower than those reported previously (6, 8) and were not sufficiently high to induce skin vasculitis in BALB/c mice. Furthermore, the predominant glomerular lesions we found were neutrophil infiltration, with no evidence of wire-loop lesions, indicating that the level of circulating cryoglobulins was not sufficient to lead to visible deposits in glomerular capillary lumens or beneath the endothelium. The role of complement in our study seemed to be independent of the type of cryoglobulinemia, because C57BL/6 mice (type I cryoglobulinemia) and BALB/c mice (type II cryoglobulinemia) developed similar glomerular lesions, and the results in complement-deficient mice of both strains were similar. The neutrophil influx observed in this study was likely to be C5 dependent, because C5-deficient mice had the lowest numbers of glomerular neutrophils detected. This would be consistent with the fact that C5a, the major cleavage product of C5, is well known to be an important chemoattactant for neutrophils. Although not excluded, it is unlikely that the membrane attack complex played a major role in this model of cryoglobulinemia, because CD59a-deficient mice had similar disease expression to wild-type controls.

Complement activation by cryoglobulin occurred via both the classical as well as the alternative pathway of complement, because in C1q- and factor B-deficient mice, similar amounts of glomerular C3 deposition were found. Such an activation of both major pathways of complement through cryoglobulins has previously been described in humans, although cryoglobulins mostly led to a consumption of the classical pathway components (19). Interestingly, in our model, although both classical and alternative pathways were activated, and both similarly contributed to C3 deposition, C5 cleavage seemed to be dependent mainly on the alternative pathway, because only factor B-deficient mice showed a reduced glomerular neutrophil influx. This finding also indicated that there was no direct role for C3 in the glomerular inflammation.

In this context, it is of note that Muhlfeld et al. (20) have previously investigated the role of Crry, a rodent membrane-bound regulator that blocks the classical and alternative pathways of complement activation, in a transgenic mouse model of mixed cryoglobulinemia (type III). In this model, overexpression of Crry did not prevent cryoglobulin-associated membranoproliferative glomerulonephritis. However, in contrast to our model, the major inflammatory cell type present in the affected glomeruli appeared to be macrophages, suggesting a different pathogenetic mechanism. Interestingly, Crry overexpression had no effect on glomerular C9 deposition, and therefore, it might not have been sufficient to achieve complete complement inhibition of the terminal pathway, including cleavage of C5.

Complement has been shown to have a protective role mediated by C1q in a model of accelerated nephrotoxic nephritis (3). However, in this model of immune complex glomerulonephritis, complement activation was deleterious, and the effect was mediated mainly by the alternative pathway. Therefore, the present study underlines the pathogenic differences between immune complex deposition and glomerulonephritis induced by anti-glomerular Abs. Furthermore, the results support the view of a dual role of complement, able to mediate protective or detrimental effects depending on the underlying pathogenic mechanisms.

More difficult to explain are the small, but consistent, differences between C3- and factor B-deficient mice. Whereas deficiency of either C3 or factor B led to markedly reduced numbers of glomerular neutrophils, only factor B-deficient mice also had significantly lower serum creatinine concentrations. Although experimental variations in a relatively small number of mice may account for these differences, it can be also postulated that in the absence of factor B, a possible beneficial effect of the classical pathway in solubilizing the immune complexes is still present, but this activity is lost in C3-deficient mice (21).

Although our findings were obtained in a model of cryoglobulinemia, they might also be of relevance for understanding other types of immune complex glomerulonephritis mediated by the deposition of Igs not specific for glomerular Ags. In particular, the glomerular disease induced by the 6-19 mAb in pristane-pretreated mice shared several features of severe lupus nephritis (9). In this respect the present study would also support the view that there is an important role of complement in the pathogenesis of immune complex-mediated glomerulonephritis. This is of particular interest, because mAbs preventing the cleavage of C5 are now available for therapy in humans (22).

We thank all the staff in the animal facility for their technical assistance.

The authors have no financial conflict of interest.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

This work was supported by the Wellcome Trust (Grant 071467). M.T. was the recipient of a fellowship from the Schweizerische Stiftung fuer Medizinisch-Biologische Stipendien (no. 1085), and L.F.-J. was the recipient of a fellowship from the Arthritis Research Campaign (United Kingdom).

4

Abbreviation used in this paper: AFI, arbitrary fluorescence intensity.

1
Schifferli, J. A., R. P. Taylor.
1989
. Physiological and pathological aspects of circulating immune complexes.
Kidney Int.
35
:
993
.-1003.
2
Unanue, E., F. J. Dixon.
1964
. Experimental glomerulonephritis. IV. Participation of complement in nephrotoxic nephritis.
J. Exp. Med.
119
:
965
.-982.
3
Robson, M. G., H. T. Cook, M. Botto, P. R. Taylor, N. Busso, R. Salvi, C. D. Pusey, M. J. Walport, K. A. Davies.
2001
. Accelerated nephrotoxic nephritis is exacerbated in C1q-deficient mice.
J. Immunol.
166
:
6820
.-6828.
4
Brouet, J. C., J. P. Clauvel, F. Danon, M. Klein, M. Seligmann.
1974
. Biologic and clinical significance of cryoglobulins: a report of 86 cases.
Am. J. Med.
57
:
775
.-788.
5
Gyotoku, Y., M. Abdelmoula, F. Spertini, S. Izui, P. H. Lambert.
1987
. Cryoglobulinemia induced by monoclonal immunoglobulin G rheumatoid factors derived from autoimmune MRL/MpJ-lpr/lpr mice.
J. Immunol.
138
:
3785
.-3792.
6
Lemoine, R., T. Berney, T. Shibata, T. Fulpius, Y. Gyotoku, H. Shimada, S. Sawada, S. Izui.
1992
. Induction of “wire-loop” lesions by murine monoclonal IgG3 cryoglobulins.
Kidney Int.
41
:
65
.-72.
7
Churg, J..
1995
.
Renal Disease: Classification and Atlas of Glomerular Diseases
Igaku-Shoin, New York, Tokyo.
8
Reininger, L., T. Berney, T. Shibata, F. Spertini, R. Merino, S. Izui.
1990
. Cryoglobulinemia induced by a murine IgG3 rheumatoid factor: skin vasculitis and glomerulonephritis arise from distinct pathogenic mechanisms.
Proc. Natl. Acad. Sci. USA
87
:
10038
.-10042.
9
Fulpius, T., R. Lemoine, T. Berney, Y. Pastore, S. Moll, S. Izui.
1996
. Polymorphonuclear leukocytes play a key role in the generation of “wire-loop” lesions induced by a murine IgG3 rheumatoid factor.
Kidney Int.
49
:
647
.-655.
10
Watanabe, N., B. Akikusa, S. Y. Park, H. Ohno, L. Fossati, G. Vecchietti, J. E. Gessner, R. E. Schmidt, J. S. Verbeek, B. Ryffel, et al
1999
. Mast cells induce autoantibody-mediated vasculitis syndrome through tumor necrosis factor production upon triggering Fcγ receptors.
Blood
94
:
3855
.-3863.
11
Izui, S., T. Fulpius, L. Reininger, Y. Pastore, T. Kobayakawa.
1998
. Role of neutrophils in murine cryoglobulinemia.
Inflamm. Res.
4
:(Suppl. 3):
S145
.-S150.
12
Botto, M., C. Dell’Agnola, A. E. Bygrave, E. M. Thompson, H. T. Cook, F. Petry, M. Loos, P. P. Pandolfi, M. J. Walport.
1998
. Homozygous C1q deficiency causes glomerulonephritis associated with multiple apoptotic bodies.
Nat. Genet.
19
:
56
.-59.
13
Fischer, M. B., M. Ma, S. Goerg, X. Zhou, J. Xia, O. Finco, S. Han, G. Kelsoe, R. G. Howard, T. L. Rothstein, et al
1996
. Regulation of the B cell response to T-dependent antigens by classical pathway complement.
J. Immunol.
157
:
549
.-556.
14
Wessels, M. R., P. Butko, M. Ma, H. B. Warren, A. L. Lage, M. C. Carroll.
1995
. Studies of group B streptococcal infection in mice deficient in complement component C3 or C4 demonstrate an essential role for complement in both innate and acquired immunity.
Proc. Natl. Acad. Sci. USA
92
:
11490
.-11494.
15
Berney, T., T. Fulpius, T. Shibata, L. Reininger, J. Van Snick, H. Shan, M. Weigert, A. Marshak-Rothstein, S. Izui.
1992
. Selective pathogenicity of murine rheumatoid factors of the cryoprecipitable IgG3 subclass.
Int. Immunol.
4
:
93
.-99.
16
Mizuochi, T., Y. Pastore, K. Shikata, A. Kuroki, S. Kikuchi, T. Fulpius, M. Nakata, L. Fossati-Jimack, L. Reininger, M. Matsushita, et al
2001
. Role of galactosylation in the renal pathogenicity of murine immunoglobulin G3 monoclonal cryoglobulins.
Blood
97
:
3537
.-3543.
17
Kikuchi, S., Y. Pastore, L. Fossati-Jimack, A. Kuroki, H. Yoshida, T. Fulpius, K. Araki, S. Takahashi, R. Lemoine, L. Reininger, et al
2002
. A transgenic mouse model of autoimmune glomerulonephritis and necrotizing arteritis associated with cryoglobulinemia.
J. Immunol.
169
:
4644
.-4650.
18
Yu, J. X., B. M. Bradt, N. R. Cooper.
2002
. Constitutive expression of proinflammatory complement components by subsets of neurons in the central nervous system.
J. Neuroimmunol.
123
:
91
.-101.
19
Adu, D., D. G. Williams.
1984
. Complement activating cryoglobulins in the nephritis of systemic lupus erythematosus.
Clin. Exp. Immunol.
55
:
495
.-501.
20
Muhlfeld, A. S., S. Segerer, K. Hudkins, A. G. Farr, L. Bao, D. Kraus, V. M. Holers, R. J. Quigg, C. E. Alpers.
2004
. Overexpression of complement inhibitor Crry does not prevent cryoglobulin-associated membranoproliferative glomerulonephritis.
Kidney Int.
65
:
1214
.-1223.
21
Schifferli, J. A., G. Steiger, G. Hauptmann, P. J. Spaeth, A. G. Sjoholm.
1985
. Formation of soluble immune complexes by complement in sera of patients with various hypocomplementemic states: difference between inhibition of immune precipitation and solubilization.
J. Clin. Invest.
76
:
2127
.-2133.
22
Fitch, J. C., S. Rollins, L. Matis, B. Alford, S. Aranki, C. D. Collard, M. Dewar, J. Elefteriades, R. Hines, G. Kopf, et al
1999
. Pharmacology and biological efficacy of a recombinant, humanized, single-chain antibody C5 complement inhibitor in patients undergoing coronary artery bypass graft surgery with cardiopulmonary bypass.
Circulation
100
:
2499
.-2506.