Modulation of Renal Disease in MRL/lpr Mice Genetically Deficient in the Alternative Complement Pathway Factor B¹

Activation of the alternative complement pathway is a first line of defense against microorganisms in the absence of specific complement-fixing Ab (1, 2, 3). Complement factor B (Bf)³ is one of the proteins required for activation of the alternative pathway. During activation of the alternative pathway, Bf binds to C3b on a pathogen surface and is cleaved to Ba and Bb by factor D. Then, properdin binds to the C3bBb complex and stabilizes it. C3bBb complex is a C3 convertase and leads to the formation of more molecules of C3b, resulting in activation of additional Bf and setting up an amplification loop for activation of C3 (1, 2, 3). Formation of the C5 convertase C3bBbC3b follows with activation of terminal complement components on the pathogen surface. This process leads to opsonization and direct damage of the pathogen (1, 2, 3). Factor B is encoded by a single gene in the MHC class III gene cluster immediately 3′ to the C2 gene; C2 is the functional homologue of Bf in the classical complement pathway (4, 5).

Factor B is produced by hepatocytes (6), phagocytes (7, 8), fibroblasts (9), epithelial cells (10), and endothelial cells (11). Bf is an acute phase reactant; therefore, serum levels of Bf increase during inflammation stimulated by cytokines, growth factors, and bacterial products (1, 5). cis elements responsible for the constitutive and regulated expression of Bf are located in the C2-Bf intergenic region (4). Factor B not only is a component of the alternative complement pathway but also serves as a B cell growth factor (12, 13), stimulates mononuclear cell cytotoxicity (14, 15), induces macrophage spreading (16), and solubilizes immune complexes (ICs) (17, 18, 19). Bf is found in glomeruli in immune complex glomerulonephritis (20), is produced by glomeruli during inflammation (20), and forms split products (5). High serum levels of these split products correlate with increased disease activity in systemic lupus erythematosus (5, 21). The role of factor B in disease pathogenesis, however, is unclear (21).

In contrast, the classical pathway of the complement system has a known important role in immune complex glomerulonephritis. ICs depositing in the kidney activate the classical pathway, thus eliciting inflammation and tissue destruction (1, 5, 22). Furthermore, components of the classical complement pathway are important in humoral responses to foreign Ags, immune complex clearance, B cell tolerance, and disposition of apoptotic cells (23, 24, 25, 26, 27). Genetic deficiency of C3 and therapies directed at blocking C3 activity (e.g., Crry transgenic mice) prevent glomerular inflammation and tissue damage in murine models of Ab-induced glomerulonephritis (28, 29, 30, 31, 32).

The role of the complement cascade in disease is not purely proinflammatory, however, as demonstrated by studies of specific complement deficiencies. Deficiencies of some components of the classical complement pathway (e.g., C4 and C1q) are associated with an increased incidence of lupus in humans and lupus-like disease in murine knockout strains (22, 23, 24). Development of lupus-like disease in C4- and C1q-deficient mice appears secondary to defects in B cell tolerance in the former and defects in clearance of apoptotic cells in the latter (23, 26). Thus, the complement cascade has both “protective” actions (i.e., clearance of ICs and apoptotic cells as well as maintenance of B cell tolerance) and “proinflammatory” effects (i.e., production of C5a and terminal complement components). The alternative pathway may enhance disease by amplifying activation of the classical pathway and its proinflammatory effects; however, the effect of the alternative pathway on inflammation and tissue destruction in immune complex glomerulonephritis is to this point unknown.

MRL/MpJ-Fas^lpr (MRL/lpr) mice spontaneously develop a severe systemic autoimmune disease similar to human lupus (33). As part of their systemic autoimmune disease, MRL/lpr mice produce autoantibodies (anti-DNA and rheumatoid factors (RF)), and cryoglobulins that are pathogenic in the immune complex glomerulonephritis and vasculitis that develop in these mice (33). Glomerular immune deposits in these mice contain IgG, IgM, IgA, Bf, and C3 (20, 33). Thus, MRL/lpr mice are useful murine models for studies of the pathogenesis of spontaneous immune complex-mediated diseases (20, 33, 34, 35).

To determine the role of the alternative complement pathway in autoimmune nephritis, Bf-deficient mice were backcrossed to MRL/lpr mice for four generations. By intercrossing F4 Bf heterozygous mice, we derived MRL/lpr B^−/−, B^+/−, and B^+/+ mice; serologic and clinical disease activity were measured over time. Our results indicate that B^−/− mice developed significantly less proteinuria, glomerular IgG deposition, and pathological renal disease than B^+/+ or B^+/− mice. Compared with the other two groups, significantly lower levels of serum IgG3 RF and cryoglobulins were found in B^−/− mice. Serum C3 levels were normal in the B^−/− mice compared with significantly decreased levels in the B^+/− or B^+/+ mice. These findings indicate that Bf plays a key role in the pathogenesis of glomerulonephritis in MRL/lpr mice and that the alternative pathway amplification loop has prominent effects on serum C3 levels in immune complex nephritis.

Factor B-deficient mice were derived as previously described (36). B6/129 B^+/− mice were intercrossed for four generations to yield the B^−/− mice used in the first crosses with MRL/lpr mice. Offspring of these matings were further backcrossed for three generations to MRL/lpr mice. B^+/− backcross four mice were bred to yield the mice used in this study. Eleven B^−/−, 14 B^+/+ and 10 B^+/− mice were included in this study. Mice were maintained in the animal facility of Washington University and the Strom Thurmond Biomedical Research Building at the Medical University of South Carolina under specific pathogen-free conditions. Routine screening of the mice for murine pathogens was negative.

Primers for PCR for Bf were as follows: W1398 (5′; Bf) 5′-CCGAAGCATTCCTATCCTCC-3′, W1399 (5′; Neo) 5′-CGAATGG GTGACCGCTTCC-3′, W1393 (3′; Bf) 5′-GTAGTCTTGTCTGCTTTCTCC-3′. DNA was isolated from tail snips (3- to 4-wk-old mice) using a QIAamp Tissue Kit (Qiagen, Santa Clarita, CA). PCR was performed by adding 2 μl DNA (1 μg) into a 23-μl reaction mixture containing 1.5 mM MgCl₂, 6.7 μM concentrations each of oligonucleotide mix, 10 mM concentrations each of dNTP mix, and 0.2 ml of Taq Gold. PCR conditions were as follows: 94°C for 4 min; 28 cycles of 94°C for 1 min, 68°C for 1 min, and 72°C for 2 min; 72°C for 10 min. After PCR amplification, samples were electrophoresed in a 2% agarose gel and visualized by ethidium bromide staining.

Primers for PCR for Fas were as follows: Fas A (5′-AGGTTAC AAAAGGTCACCC-3′), Fas B (5′-GATACGAAGATCCTTTCCTGTG-3′), and Fas C (5′-CAAACGCAGTCAAATCTGCTC-3′). In brief, individual genomic tail DNA was isolated and used in PCR. PCR conditions were as follows: 94°C for 10 min; 40 cycle of 94°C for 1 min, 50°C for 1 min, and 72°C for 1 min; 72°C for 10 min. After PCR amplification, samples were electrophoresed in a 2% agarose gel and visualized by ethidium bromide staining.

H-2 of the mice was determined by flow cytometric analysis of splenic mononuclear cells. Single-cell suspensions were derived and stained with FITC-labeled anti-H-2D^k (mAb KH95) or anti-H-2D^b (mAb 15-5-5, PharMingen, San Diego, CA). FITC-labeled isotype control Abs were used to eliminate nonspecific staining as a variable. Samples were analyzed on a Becton Dickinson FACStar machine, and results analyzed with CellQuest software (Becton Dickinson).

To confirm the H-2 genotype, we performed PCR RFLP analysis using the techniques of Peng and Craft (37) specifically designed to differentiate H-2^k and H-2^b. Briefly, tail DNA was amplified using the primers IAAIF 5′-GAAGACGACATTGAGGCCGACCACGTAGGC-3′ and IAAIR 5′-ATTGGTAGCTGGGGTGGAATTTGACCTCTT-3′. PCR was performed with a 12-min denaturation step of 94°C followed by cycles of 94°C for 20 s, 60°C for 1 min, and 72°C for 1 min with a final extension of 10 min at 72°C. The resultant PCR product was digested with HindIII and run on an agarose gel. Amplified DNA of the H-2^k genotype is digested whereas DNA of the H-2^b genotype is not, allowing differentiation by agarose gel electrophoresis.

Mice were placed in metabolic cages for 24-h urine collection. To prevent bacterial growth, antibiotics (ampicillin, gentamicin, and chloramphenicol) were added to collection tubes. Urinary protein excretion was determined with the Bio-Rad protein assay kit according to the manufacturer’s instructions (Bio-Rad, Hercules, CA) and reported as milligrams of protein per mouse per day.

Anti-DNA levels were measured by ELISA as previously described (38). Briefly, 96-well ELISA plates were coated with 5 μg/ml double-stranded calf thymus DNA (dsCT DNA) at 37°C overnight. The plates were then washed with PBS-0.05% Tween (PBS-T). Sera were added in serial dilutions, starting at a 1/100 dilution to each well, and incubated for 45 min at room temperature (RT). After a washing with PBS-T, HRP-conjugated goat anti-mouse IgG (γ-chain specific, Sigma, St. Louis, MO) was added and incubated for 45 min. After additional washing, substrate solution containing 3,3′,5,5′-tetramethylbenzidene (TMB, Sigma) was added in 0.1 M citrate buffer, pH 4, and 0.015% H₂O₂. After incubation for 30 min, absorption at OD₃₈₀ was determined on a Flow microtiter plate reader (Dynatech, McLean, VA). Results are shown as the OD₃₈₀ absorbance at a 1/100 dilution. dsDNA was derived by S₁ nuclease (Sigma) treatment.

Cryoglobulins were isolated from sera as previously described (39). Briefly, blood samples were placed at 37°C for 2 h. After centrifugation at 1000 rpm, supernatants were collected and placed at 4°C for 72 h. They were then centrifuged at 3000 rpm, supernatants were removed, and precipitates were washed five times with cold PBS. After being washed, they were resuspended in PBS at the same volume as the original sera. The isolated cryoglobulins were placed at 37°C for 3 h before use to ensure resolubilization.

Total IgG levels in sera or in cryoglobulins were determined by ELISA using a standard curve of known concentrations of mouse IgG. Cryoglobulins were placed at 37°C for 3 h before use. ELISA plates were coated with 1 μg/ml anti-mouse κ overnight at 4°C. Sera or cryoglobulins were added in serial dilutions starting at a 1/1000 dilution. HRP-conjugated goat anti-mouse IgG (γ-specific, Sigma) was added, followed by TMB for color development. OD₃₈₀ absorbance was measured as above. The same method was used for measurement of serum Ig isotype levels and total Ig in serum and in cryoglobulins except HRP goat anti-mouse IgG1, IgG2a, or IgG3 or HRP goat anti-mouse H & L chain were used (Southern Biotechnology, Birmingham, AL).

Circulating ICs were determined by the C1q ELISA method as previously described with minor modifications (40). ELISA plates were coated with human C1q in 0.1 M carbonate buffer (pH 9.6) and incubated for 48 h at 4°C. After a washing with PBS, PBS-1% BSA was added to each well and incubated for 2 h at RT. EDTA-treated sera samples (diluted 1/50) were added in duplicate and then incubated for 1 h at RT and then overnight at 4°C. HRP anti-mouse IgG (γ-chain specific, Sigma) was added in PBS-T. The rest of the assay was performed as described above. Aggregated human γ-globulin was used as a positive control.

RF concentrations in sera were measured by ELISA as previously described (33). ELISA plates were coated with rabbit IgG (1 μg/ml) overnight at 4°C. After a washing, sera were added in serial dilution starting at 1/100 dilution. The assay was then performed as described above with HRP-conjugated goat anti-mouse IgG (γ-chain specific, Sigma) or goat anti-mouse IgM (Sigma). For the measurement of IgG3 anti-IgG2a RF, the same method was used except ELISA plates were coated with mouse IgG2a and HRP anti-mouse IgG3 was used for detection of RF activity.

C3 levels in sera were measured by ELISA. ELISA plates (with 96 wells) were coated with goat anti-mouse C3 (Cappel, Durham, NC) and incubated overnight at 4°C. After washing and blocking with 5% BSA in PBS for 1–2 h, samples diluted 1/200, 1/1,000, 1/5,000, and 1/25,000 were added to individual wells and incubated for 1–2 h at RT. After a washing, rat anti-mouse C3 mAb (RmC11H9) was added and incubated for 1 h at RT (28). Peroxidase-conjugated anti-rat IgG was added and then incubated for 1 h at RT. 2,2′-azinobis(3-ethylbenzthiazoline-6-sulfonic acid) (1 mM) in 0.1 M citric acid buffer and 0.03% H₂O₂ was added for color development. The plates were read on an ELISA reader at OD₄₀₅.

Factor B levels were measured by ELISA. ELISA plates were coated with goat anti-human factor B IgG fraction (DiaSorin, Stillwater, MN) overnight at 4°C. Sera diluted to 1/100, 1/200, and 1/400 were added to individual wells. Biotinylated anti-human Bf Ab (Diasorin) was added followed by avidin-peroxidase. TMB was added for color development. After incubation for 30 min, OD₃₈₀ absorption was determined.

At the time of sacrifice (22 and 44 wk), the kidneys were removed. One kidney was fixed with formaldehyde, embedded in paraffin, and then sectioned before staining with hematoxylin and eosin (H&E) and periodic acid-Schiff (PAS). The slides were read and interpreted in a blinded fashion, grading the kidneys for glomerular inflammation, proliferation, crescent formation, and necrosis. Interstitial changes and vasculitis were also noted. Scores from 0 to 3 were assigned for each of these features and then added together to yield a final renal score (33). For example, glomerular inflammation was graded: 0, normal; 1, few inflammatory cells; 2, moderate inflammation; and 3, severe inflammation. The other kidney was snap frozen in liquid nitrogen and placed in OCT medium. Frozen sections 4 μm thick were stained with fluorescein-conjugated anti-mouse IgG, IgM, IgA, IgG3, or C3 (Cappel, Durham, NC). Slides were read in a blinded fashion and graded 0–3+ (1+, mild staining; 2+, moderate staining; 3+, high staining).

Statistical values were determined by using the Mann-Whitney two-tailed U test or Kruskal-Wallis test. A p value <0.05 was considered significant.

To determine the role of Bf in autoimmune nephritis, Bf knockout mice were backcrossed for four generations to MRL/lpr mice. F4 B^+/− mice were then intercrossed to yield the cohorts for these studies. The genotype of the mice was determined by PCR for Bf and Fas. The percentage of offspring in the litters derived were consistent with the predicted outcome; i.e., ∼25% of pups were either homozygous knockout (B^−/−, n = 11) or wild type (B^+/+, n = 14), whereas 50% were heterozygous for B⁻ (n = 20). Ten B^+/− mice were randomly selected as the study mice. There were no differences in spleen weight or body weight between groups at 22 or 44 wk of age (data not shown). By gross examination at the time of sacrifice, 100% of female B^+/+ mice (7 of 7) had an excoriated skin rash on the shoulder area as well as tail and ear necrosis; 1 of 7 male B^+/+ mice had ear and tail necrosis. Neither skin rash nor appendage necrosis was observed in B^−/− or B^+/− mice. These skin lesions are secondary to small vessel vasculitis (41, 42).

To assess the MHC expression by each group, H-2 expression of splenocytes was assessed by flow cytometry. The MRL/lpr Bf-deficient mice were H-2D^b; the Bf heterozygotes H-2D^{b, k}, and the wild-type mice were H-2D^k. PCR RFLP analysis of MHC class II genotype confirmed the H-2 haplotype of the mice: B^−/− mice were H-2^b; B^+/− mice were H-2^b/k; and B^+/+ mice were H-2^k.

Bf levels in sera were measured by ELISA when the mice were 22 wk of age. As shown in Table I, no Bf was detected in the sera of B^−/− mice, confirming that the phenotype of the mice matched the genotype. Serum Bf levels of B^+/+ mice were significantly higher than that of B^+/− mice (p ≤ 0.05), indicating a gene dose effect on serum Bf levels.

To assess the role of Bf in autoimmune nephritis, we measured 24-h urine protein excretion beginning at 12 wk of age. As shown in Fig. 1, there was a statistically significant difference in proteinuria comparing B^+/− mice after 20 wk of age (p ≤ 0.05) to both B^+/+ mice and B^−/− mice, indicating the development of early severe disease in the heterozygous mice. The B^+/+ mice developed increasing proteinuria after 26 wk of age that was significantly greater than that in the B^−/− mice (p ≤ 0.05). Even at 44 wk of age, B^−/− mice still had not developed proteinuria >1 mg/mouse/day (data not shown).

A group of mice were sacrificed at the age of 22 wk (5 B^−/−, 7 B^+/+, and 3 B^+/− mice), and pathologic renal scores were determined using a previously published scale for glomerular inflammation (33). The remaining B^+/+ and B^−/− mice were sacrificed at 44 wk (all B^+/− mice had died or were sacrificed before 25 wk). As shown in Fig. 2,A, Bf-deficient mice had lower renal scores than the mice in the other groups at both 22 and 44 wk. The renal abnormalities in the MRL/lpr B^−/− mice consisted of focal glomerular hypercellularity without significant infiltrate of inflammatory cells (Fig. 2,B). Glomerular wire loop lesions, immune complex deposition, and glomerular hyalinization, as determined by intensity of PAS staining, were minimally present in the B^−/− mice (Fig. 2,C). Renal findings in the MRL/lpr B^+/− (data not shown) and B^+/+ mice included diffuse glomerular hypercellularity, wire loop lesions, mesangial expansion, crescent formation, fibrinoid necrosis, glomerular hyalinization, and infiltration of inflammatory cells (Fig. 2, D and E). The difference in renal scores in the B^−/− compared with the other two groups at 22 wk was not statistically significant (p = 0.13) due to severe disease in one male B^−/− mouse with a renal score similar to that seen in the other two groups. We included the renal score of this mouse in the overall score and statistical analysis even though it exceeded the mean of the scores of the other B^−/− mice by 6 SDs. To ensure that this mouse had been genotyped correctly, spleen DNA was amplified and results of the PCR amplification confirmed the B^−/− genotype of the mouse. If the score for this mouse is excluded, the difference in renal score between the B^−/− mice and either of the other two groups was significant at p ≤ 0.01. The difference in renal scores between the B^−/− mice and B^+/+ mice was significant at 44 wk (p = 0.04).

To determine the effects of Bf genotypes on glomerular IC deposition, immunofluorescence analysis was performed. Frozen sections were stained with fluorescein-conjugated anti-mouse C3, IgG, IgG3, IgM, and IgA. C3, IgM, and IgA deposition were similar in all groups with no apparent qualitative or quantitative differences (data not shown). Glomerular IgG3 deposition was minimal in all three groups. There was, however, significantly less IgG deposition in B^−/− than in B^+/+ mice (p ≤ 0.05, Table II). IgG deposition in the kidneys of B^−/− mice was also less than in the B^+/− mice; the difference, however, did not reach statistical significance (Table II).

Medium to large vessel vasculitis is a pathologic feature of renal disease in MRL/lpr mice (41). At the age of 22 wk, of the 5 B^−/− mice examined, only the male B^−/− mouse with severe renal disease had vasculitis (1 of 5). Of female B^+/+ mice 50% (2 of 4) had vasculitis as illustrated in Fig. 2 F, whereas no vasculitis was present in male B^+/+ mice (0 of 3). In the B^+/− mice, 1 of 3 of the mice had vasculitis; all 3 B^+/− mice studied were females. At the age of 44 wk, 100% of female B^+/+ mice had vasculitis (3 of 3). There was no vasculitis in male B^+/+ mice (0 of 4) or vasculitis observed in the 6 B^−/− mice (3 males and 3 females). These data indicate a greater incidence of vasculitis in the B^+/+ mice than in the B^−/− mice (overall 1 of 11 B^−/− mice had vasculitis compared with 5 of 14 B^+/+ mice). This difference in incidence of vasculitis was especially evident in female mice where 0 of 6 female B^−/− mice had vasculitis compared with 5 of 7 female B^+/+ mice (p ≤ 0.01). The presence of renal vasculitis in the female B^+/+ mice paralleled the vasculitis induced ear and tail necrosis present also primarily in the female B^+/+ mice.

To investigate possible mechanisms for the modulation of renal disease in the B^−/− mice, we measured levels of autoantibodies implicated in disease pathogenesis. Anti-DNA Abs are pathogenic in the autoimmune nephritis that develops in MRL/lpr mice (33). Anti-dsCT DNA Abs in serum were measured by ELISA. As shown in Fig. 3, serum anti-dsCT DNA Ab levels were lower at 16 wk of age in B^−/− mice than in the other two groups (p = 0.02). However, by 20 wk and thereafter, the serum anti-DNA levels were not significantly different between the groups out to 44 wk of age. Serum anti-glomerular binding activity, as measured by ELISA, paralleled anti-DNA activity with lower serum levels at 16 wk in the B^−/− mice that equilibrated between the groups over time (data not shown).

To assess the effect of factor B on the production of other autoantibodies in the B^−/− mice, serum RF activity was measured by ELISA. There were no detectable differences between the groups in production of total IgM RF or IgG RF (data not shown). However, as presented in Fig. 4, significantly lower levels of IgG3 anti-IgG2a RF, a specificity linked with vasculitis in these mice, were present in the sera of B^−/− mice at the age of 20 wk than in sera of the other 2 groups (41, 42). These differences were maintained through 44 wk.

MRL/lpr renal disease and vasculitis has been linked with IgG3 cryoglobulin production (41, 42, 43, 44). To determine whether the Bf genotype affected Ig or cryoglobulin production, total serum Ig levels and IgG3 cryoglobulin levels were measured by ELISA. Total IgG levels were similar at all time points in all three groups (data not shown); however, significantly higher levels of serum IgG3 and IgG3 cryoglobulins were present in the B^+/− and B^+/+ mice compared with the B^−/− mice (Table III, Fig. 5). Serum levels of all other IgG isotypes were not statistically different between the 3 groups (data not shown).

To determine the effects of a lack of Bf on serum C3 levels, C3 levels were determined by ELISA at the ages of 16, 20, and 24 wk in all three groups. As shown in Table IV, serum C3 levels in the B^−/− mice were equivalent to levels in normal C57BL6 mice and were significantly higher than serum C3 levels in the other two groups. Levels of serum C3 in the B^−/− mice decreased over time, suggesting some activation of C3, but were still significantly higher than the B^+/+ mice even at 44 wk of age (data not shown). These data imply a significant role for Bf and the alternative pathway amplification loop in the serum C3 levels in this murine model of immune complex renal disease.

To assay for effects of Bf on clearance of circulating ICs, circulating ICs were measured by ELISA using the C1q binding assay. There were no differences in sera circulating ICs between groups at any time point (data not shown).

To provide insight into the role of the alternative pathway in lupus, we backcrossed Bf-deficient mice to MRL/lpr mice for four generations and then intercrossed B^+/− mice. Our results indicate that MRL/lpr B^−/− mice had significantly less vasculitis, less proteinuria, lower pathologic renal scores, and less glomerular IgG deposition with lower levels of serum IgG3 and IgG3 cryoglobulins than Bf-producing littermates. Furthermore, MRL/lpr B^−/− mice had normal serum C3 levels compared with significantly decreased levels in B^+/+ and B^+/− mice. Thus, these studies demonstrate a key role for Bf in the pathogenesis of autoimmune renal disease and vasculitis in MRL/lpr mice. In addition, these results demonstrate a role of Bf and the amplification loop in the production, activation, or consumption of C3 in this disease model.

The mechanism by which Bf deficiency abrogates disease may involve effects on the inflammatory cascade through decreased C3 activation or direct B cell effects or a combination of both. At this time, we cannot determine which of these effects is responsible for blocking disease development in the B^−/− mice, although differentiating these effects is experimentally approachable. Considering the effect of Bf deficiency on C3 activation alone, using a simplified interpretation of a complex biologic system, C3 has both “protective antiinflammatory” functions (i.e., transporting or solubilizing ICs) and “destructive proinflammatory” functions (i.e., promoting local deposition of terminal complement components and release of inflammatory mediators, i.e., C5a). Serum C3 levels represent a complex interplay of C3 production, activation, and consumption. Although we postulate that the effects of factor B deficiency on serum C3 levels are through decreased C3 activation and consumption, it is feasible that factor B deficiency may effect C3 production. The finding of decreased serum C3 activation in MRL/lpr Bf-deficient mice is consistent with similar findings noted in nonautoimmune Bf-deficient mice (36).

In lupus nephritis, the alternative pathway may enhance local tissue damage in disease via the amplification loop in which C3b, activated by IC deposition, initiates the alternative pathway feeding back on and accentuating C3 activation. In this case, the alternative pathway promotes proinflammatory complement functions including the generation of the chemotactic factor C5a. In B^−/− mice, the proinflammatory functions of C3 may be diminished due to the lack of C3 activation via the amplification loop. For example, Bf deficiency may block local intrarenal inflammation by modulating levels of specific components of the complement cascade such as C5a. The lack of Bf appears to have inhibited the production of such chemotactic factors as there was minimal to no inflammatory infiltrate in the kidneys of 10 of 11 B^−/− mice; inflammatory infiltrates were present in all of the kidneys from the B^+/− and B^+/+ mice. This lack of inflammatory mediators likely also limited the development of vasculitis in the B^−/− mice.

The early decreased levels of anti-DNA Abs and the decreased deposition of IgG in the kidneys of B-deficient mice may reflect alterations in B cell function and autoantibody production in the B^−/− mice. Although the overall effects of Bf on B cell function are controversial, previous studies reported stimulation of B cell proliferation and differentiation by Bf (13). In nonautoimmune Bf-deficient mice, however, no alterations in B cell proliferation or differentiation or responses to T-dependent or T-independent Ags were noted (36). In the MRL/lpr Bf-deficient mice there appears to be a possible linkage between disease severity and the B cell production of IgG3-containing cryoglobulins. Recent studies by two different groups demonstrated that IgG3 cryoglobulins are pathogenic in vasculitis and glomerulonephritis in MRL/lpr mice (39, 40, 41, 42, 43). An IgG3 mAb with anti-IgG2a RF activity derived from a MRL/lpr mouse induced skin vasculitis and glomerulonephritis when injected into normal mice (42). IgG3 cryoglobulins also induced “wire loop” lesions in glomeruli by direct glomerular deposition independent of IC deposition and complement activation (43, 44).

The low concentrations of IgG3 cryoglobulins appear directly linked to low serum levels of total IgG3 in the MRL/lpr B^−/− mice and suggest modulation of B cell isotype switching in Bf-deficient mice. That serum levels of IgG2a in the B^−/− mice were equivalent to the other two groups indicates that there was no overall defect in isotype switching. Serum IFN-γ levels were similar in all three genotypes (data not shown); thus, it is unlikely that modulation of IFN-γ production alone was a mechanism for the decreased sera levels of IgG3 in the B^−/− mice. Studies of the B^−/− genotype on a “normal” strain background found no differences in B cell number, germinal center formation, or IgG production (36). C3-deficient mice, however, do have defects in B cell function including decreased isotype switching and blunted responses to T-dependent Ags such as DNA and IgG (22, 23). An indirect effect of Bf deficiency on B cell function via decreased C3 activation may only become evident in the setting of an activated or abnormal immune system such as occurs in MRL/lpr mice.

Independent of effects on inflammation, complement components may also modulate autoimmune disease through indirect effects on B cells. As recently demonstrated by the Shlomchik group, B cells, independent of Ig production, play a key role in disease development in MRL/lpr mice presumably via Ag presentation (45). Complement opsonization is involved in the presentation of immune complexes by resting B cells. Thus, the lack of C3 activation in the B^−/− mice may alter B cell-mediated Ag presentation resulting in a diminished autoimmune response. CD 21 can directly bind immune complexes and present them to T cells and may also augment MHC presentation of Ag; decreased complement activation may also limit this alternative mechanism of Ag presentation (46, 47).

An alternative explanation to account for disease amelioration and decreased C3 consumption in the B^−/− mice is that IgA immune complexes did not activate the alternative pathway in the Bf-deficient mice (1). In Bf-producing mice, IgA ICs activate the alternative pathway leading to C3 activation (1). IgA ICs are present in the kidneys of MRL/lpr mice and patients with lupus nephritis, although their contribution to disease pathogenesis is unclear (5). Equivalent deposition of IgA in the kidneys was present in all three groups of MRL/lpr mice; however, in the B^−/− mice, this deposition would not have activated the alternative complement pathway or activated C3 due to the lack of Bf. Additional studies are needed to determine the contributions of IgA IC-mediated complement activation on disease modulation in MRL/lpr mice.

An unexplained finding in these studies is the lack of a Bf dose response in many of the parameters measured; i.e., Bf heterozygous mice had an earlier onset of disease and greater proteinuria than wild-type mice. The concept of protective vs proinflammatory effects of complement in disease pathogenesis may explain why B^+/− mice had worse disease than B^+/+ mice (i.e., immune complex solubilization was diminished, but production of inflammatory mediators was not). Bf levels in the B^+/− mice appear sufficient to activate C3 and promote inflammation as evidenced by the presence of inflammatory cells in the glomeruli and decreased serum levels of C3. The presence of IgG3 cryoglobulins in the sera of the B^+/− mice suggests that the lower concentrations of Bf in these mice were sufficient for IgG3 and IgG3 cryoglobulin production. In contrast, there may not have been sufficient complement activation in the B^+/− mice for adequate solubilization and clearance of immune complexes, although we have no experimental evidence indicating such a defect. We do believe, however, that the increased disease in B^+/− mice is biologically relevant because we also found similar early severe renal disease in MRL/lpr C3^+/− mice compared with C3^+/+ and C3^−/− mice (48). The similar findings of accelerated disease in both MRL/lpr B^+/− mice and C3^+/− mice suggest that low levels of these complement factors may be more pathogenic in IC disease than either completely absent or normal levels. The similarities of disease expression in the C3 and Bf heterozygous mice also suggest that the mechanism for accelerated disease in the B^+/− mice is mediated via effects on C3.

As in all gene inactivation studies, careful consideration must be given to alternative mechanisms for the biological effects seen in the knockout mice. For example, the genes for Bf and C2 are contiguous on mouse chromosome 17 (4). Although we did not specifically measure C2 levels in the MRL/lpr B^−/− mice, studies of the B^−/− founder mice for our initial breeding indicated normal C2 levels and function (36). A more significant consideration in the mice in this study is the potential effect of H-2 on disease. Although the mice in this study are ∼97% homogeneous for the MRL genetic background, Bf is part of the MHC class III locus and is thus linked to the rest of the MHC locus. Because these mice were selected based on Bf genotype, it is likely that a significant portion of chromosome 17, including the MHC locus, was linked with the Bf genotype and remains of 129 genetic derivation.

H-2 affects Ag handling and peptide presentation; thus, H-2 differences could affect production of autoantibodies through alterations in Ag presentation. By spleen cell flow cytometric analysis and PCR RFLP analysis of tail DNA, the B^−/− mice were H-2^b. The H-2^b present in the B^−/− mice was derived from the original knockout construct in 129 mice. The B^+/+ mice were H-2^k (the H-2 of MRL/lpr mice) and the heterozygous mice were H-2^b/k. H-2 and H-2 heterozygosity are known to play a major role in disease in some strains of lupus-prone mice (e.g., NZB/NZW mice) (34, 35). Such an effect of H-2 on MRL/lpr disease, however, has not been reported. In genetic linkage studies in MRL/lpr mice, H-2 is not linked with anti-DNA production, anti-glomerular Ab binding, or incidence or severity of renal disease in H-2^k × H-2^b crosses involving MRL/lpr mice or further backcrosses to MRL/lpr where H-2^b/b mice were derived (34, 35). The only linkage between disease and H-2 in MRL (H-2^b × H-2^k) crosses and backcrosses is that MRL/lpr H-2^b mice are higher producers of IgM and IgG RFs than MRL H-2^k mice (34). We did not observe this effect in the MRL/lpr B^−/− mice (data not shown). The H-2^b in the factor B knockout mice was derived from 129 mice. Although the H-2^b of C57BL/6 mice does not appear to affect disease expression in MRL/lpr mice, it is conceivable there are differences in the H-2^b of 129 mice and factor B mice that could affect immune function. We are unaware of any reports of such differences. We cannot definitively rule out a contribution of genetic elements other than factor B to the phenotypic differences seen in the study mice. We believe, however, that the use of littermates as controls and the current data regarding effects of MHC on disease in MRL/lpr mice suggests the contributions of these other genetic elements is likely minimal.

In summary, our results suggest that: 1) Bf has an important role in the development of glomerulonephritis and vasculitis in MRL/lpr mice through direct or indirect effects on the production of specific autoantibodies, cryoglobulins and inflammatory mediators; 2) Bf and the alternative pathway amplification loop play an important role in C3 production, activation, or consumption in this model of immune complex renal disease. MRL/lpr B^−/− mice thus provide a new and useful model for dissecting the role of the alternative pathway in IC- mediated diseases. Furthermore, these results suggest that Bf may be a novel specific target for therapy in IC-mediated diseases.

We thank Dr. M Matsumoto who made the original factor B knockout construct and derived the original factor B-deficient mice.