Title: Mesangial immune complex glomerulonephritis due to complement factor D deficiency
Abstract: Complement factor D is a serine protease essential for the activation of the alternative pathway and is expressed in the kidney, adipocytes, and macrophages. Factor D is found at relatively high levels in glomeruli suggesting that this component of the complement cascade could influence renal pathophysiology. In this study, we utilize mice with a targeted deletion of the activating complement factor D gene and compare these results to mice with targeted deletion of the inhibitory complement factor H gene. Eight-month-old mice with a deleted factor D gene spontaneously develop albuminuria and have reduced creatinine clearance due to mesangial immune complex glomerulonephritis. These mesangial deposits contain C3 and IgM. In contrast to the mesangial location of the immune deposits in the factor D-deficient mice, age-matched factor H-deficient mice develop immune deposits along the glomerular capillary wall. Our observations suggest that complement factor D or alternative pathway activation is needed to prevent spontaneous accumulation of C3 and IgM deposits within the mesangium. Our studies show that the complement factor D gene knockout mice are a novel model of spontaneous mesangial immune complex glomerulonephritis. Complement factor D is a serine protease essential for the activation of the alternative pathway and is expressed in the kidney, adipocytes, and macrophages. Factor D is found at relatively high levels in glomeruli suggesting that this component of the complement cascade could influence renal pathophysiology. In this study, we utilize mice with a targeted deletion of the activating complement factor D gene and compare these results to mice with targeted deletion of the inhibitory complement factor H gene. Eight-month-old mice with a deleted factor D gene spontaneously develop albuminuria and have reduced creatinine clearance due to mesangial immune complex glomerulonephritis. These mesangial deposits contain C3 and IgM. In contrast to the mesangial location of the immune deposits in the factor D-deficient mice, age-matched factor H-deficient mice develop immune deposits along the glomerular capillary wall. Our observations suggest that complement factor D or alternative pathway activation is needed to prevent spontaneous accumulation of C3 and IgM deposits within the mesangium. Our studies show that the complement factor D gene knockout mice are a novel model of spontaneous mesangial immune complex glomerulonephritis. The complement system can be activated by the classical, lectin, or alternative pathways to generate 'convertases', enzyme complexes that cleave C3 and C5, leading to the formation of biologically active proteins, including anaphylatoxins (C3a and C5a), opsonic fragments (C4b, C3b, and iC3b) and the membrane attack complex (C5b-9).1.Walport M.J. Complement – first of two parts.N Engl J Med. 2001; 344: 1058-1066Crossref PubMed Scopus (2217) Google Scholar Together these activated proteins mediate the biologically important functions of complement including host defense (opsonization, leukocyte activation, and target cell lysis), immune complex clearance, and an accessory role in antibody production, through its role as a natural adjuvant. Factor D is a highly specific serine protease that is essential for alternative pathway activation.2.Lesavre P. Muller-Eberhard H. Mechanism of action of factor D of the alternative complement pathway.J Exp Med. 1978; 148: 1498-1509Crossref PubMed Scopus (118) Google Scholar Activated C3, formed either spontaneously through hydrolysis of the internal thioester bond (C3i) or enzymatically by the C3 convertase complexes (C3b) interacts with the complement protein factor B. Factor B bound to activated C3 is cleaved by factor D releasing a 20 kDa activation fragment (Ba).3.Gotze O. Muller-Eberhard H.J. The C3-activator system: an alternate pathway of complement activation.J Exp Med. 1971; 134: 90-108Crossref PubMed Scopus (579) Google Scholar The larger factor B fragment (Bb) remains bound to activated C3 and this complex represents the alternative pathway C3 convertase (C3iBb, C3bBb). Thus, in the absence of factor D cleavage of factor B bound to activated C3 does not occur, preventing alternative pathway C3 convertase formation. In contrast, factor H is the major fluid-phase inhibitory protein of alternative pathway activation. Its critical role in downregulating alternative pathway activation is illustrated by the observation that individuals with genetic deficiency of factor H develop secondary markedly reduced plasma C3 levels due to C3 consumption as a consequence of uncontrolled spontaneous alternative pathway activation. Both thrombotic microangiopathy and membranoproliferative glomerulonephritis type II (MPGN type II, also termed dense deposit disease) have been reported in association with factor H dysfunction.4.Rodriguez de Cordoba S. Esparza-Gordillo J. Goicoechea de Jorge E. et al.The human complement factor H: functional roles, genetic variations and disease associations.Mol Immunol. 2004; 41: 355x-367xCrossref PubMed Scopus (463) Google Scholar Notably, complete deficiency of this protein is usually associated with MPGN type II, whereas thrombotic microangiopathy is generally associated with mutations that specifically target surface recognition domains of the protein.4.Rodriguez de Cordoba S. Esparza-Gordillo J. Goicoechea de Jorge E. et al.The human complement factor H: functional roles, genetic variations and disease associations.Mol Immunol. 2004; 41: 355x-367xCrossref PubMed Scopus (463) Google Scholar Factor D is expressed at relatively high levels in glomeruli,5.Song D. Zhou W. Sheerin S. Sacks S. Compartmental localization of complement component transcripts in the normal human kidney.Nephron. 1998; 78: 15-22Crossref PubMed Scopus (32) Google Scholar suggesting that local synthesis of this complement component may play a role in renal function. To determine whether factor D has a physiological relevant role within the kidney, we have studied the spontaneous renal phenotype of 8-month-old gene-targeted factor D-deficient mice (Cfd−/−,6.Xu Y. Ma M. Ippolito G.C. et al.Complement activation in factor D-deficient mice.Proc Natl Acad Sci. 2001; 98: 14577-14582Crossref PubMed Scopus (127) Google Scholar). We compared the renal pathology of these animals with the spontaneous pathology that develops in factor H-deficient mice (Cfh−/−,7.Pickering M.C. Cook H.T. Warren J. et al.Uncontrolled C3 activation causes membranoproliferative glomerulonephritis in mice deficient in complement factor H.Nat Genet. 2002; 31: 424-428Crossref PubMed Scopus (386) Google Scholar). Thus, analysis of these two mutant strains enabled us to contrast the spontaneous renal histological changes that develop when alternative pathway activation is either prevented (Cfd−/−) or uncontrolled (Cfh−/−). Our data demonstrated that C57BL/6 Cfd−/− mice developed extensive immune deposits within the mesangium while the spontaneous immune deposits that develop in the Cfh−/− animals were located along the capillary walls. Polymerase chain reaction (PCR) amplification in wild-type (WT) controls generated the expected 500 bp amplified product, whereas in both Cfh−/− and Cfd−/− mutant mice a 400 bp product was obtained (Figure 1). There was no expression of the factor H gene in the kidneys or liver of the Cfh−/− mutants, and renal expression of Factor D was absent in the Cfd−/− mutants (Figure 2). In both homozygous mutants, expression of C3 in the kidney and liver was increased as compared with WT controls (Figure 2).Figure 2RT–PCR of WT, Cfh−/−, and Cfd−/− kidney and liver tissues. Glyceraldehyde 3-phosphate was used as a control.View Large Image Figure ViewerDownload (PPT) The findings are summarized in Table 1. Both Cfd−/− and Cfh−/− mutant mice had decreased creatinine clearance and significant albuminuria when compared with WT controls (P<0.05), the decline in renal function was greater in the Cfh−/− mutants.Table 1Renal function analysis of Cfh−/−, Cfd−/−, and WT miceMouseWeight (g)24 h urine volume (ml)Serum creatinine (mg/dl)Urine creatinine (mg/dl)Urine albumin (mg/dl)Creatinine clearance (ml/24 h)Cfh−/− (n=6)30.331.000.64±0.084.89±0.87597±94 (P=0.0004*P<0.05 Cfh or Cfd vs WT.)7.35±1.08 (P=0.0005*P<0.05 Cfh or Cfd vs WT.)Cfd−/− (n=5)29.61.000.74±0.215.86±0.90349±94 (P=0.019*P<0.05 Cfh or Cfd vs WT.)8.58±3.07 (P=0.023*P<0.05 Cfh or Cfd vs WT.)WT (n=5)301.000.51±0.226.74±0.66153±5817.33±5.33SEM, standard error of mean, WT, wild-type.Values are in mean±SEM.* P<0.05 Cfh or Cfd vs WT. Open table in a new tab SEM, standard error of mean, WT, wild-type. Values are in mean±SEM. Glomeruli in Cfd−/− mice show mild mesangial expansion with an increase in mesangial matrix and mild increase in cellularity (Figure 3b) as compared with WT mice (Figure 3a). The increase in mesangial cellularity was due to mild increase in mononuclear cells. On average, there were 31 cells (±3.0) per glomerulus in the Cfd−/− mice, whereas WT mice contained 25 cells (±1.5) per glomerulus. The mesangial cells in the Cfd−/− mice stained for predominantly for smooth muscle actin (Figure 4b), whereas they were negative for F4/80 (Figure 4e), indicating that the increase in cellularity was due to mesangial cell proliferation and not infiltrating mononuclear leukocytes. Glomerular capillaries did not show significant changes, and in particular, endocapillary proliferation or formation of double contours along the capillary walls was absent. Segmental and global glomerulosclerosis was not present, and neither was there significant tubulointerstitial disease. Isolated tubules contained red blood cells. Glomeruli in Cfh−/− mice were hypercellular (Figure 3c). The mesangium was expanded with increased cellularity primarily due to mononuclear cells. A few glomerular capillaries showed endocapillary proliferation due to mononuclear cells, and glomerular tufts had a distinctly lobular appearance. On average, there were 40 cells (±5.2) per glomerulus. Again, the mesangial cells stained for smooth muscle actin (Figure 4c), whereas they were negative for F4/80 (Figure 4f), indicating that they were mesangial cells and not infiltrating mononuclear leukocytes. Glomerular capillary walls were thickened and new basement membrane formation could be seen resulting in double contour formation. Segmental sclerosis was also noted. Significant tubulointerstitial disease was not present (Table 2). WT, wild-type. Glomeruli in Cfd−/− mice showed marked granular staining for C3 (3+) and IgM (3+) (Figure 5c and d) in the mesangium compared with WT mice (Figure 5a and b). There was also mild mesangial staining for IgG (1–2+) and IgA (+/-), but virtually no C3 or other immune deposits along the glomerular capillary walls. In contrast, glomeruli in Cfh−/− mice showed marked granular staining for C3 and IgM predominantly along glomerular capillary walls (Figure 5e and f). There was also mild positive staining for IgG (1–2+) and IgA (trace) along the capillary walls. WT mice showed normal appearing mesangium and glomerular capillary walls. Electron-dense deposits were not present in the mesangium or along the capillary walls (Figure 6a). Ultrastructural examination of glomeruli of Cfd−/− mice showed mesangial expansion with an accumulation of electron-dense deposits (Figure 6b–d), which were mesangial and paramesangial in location. In comparison, Cfh−/− mice showed changes of membranoproliferative glomerulonephritis with subendothelial expansion by basement membrane material, cellular debris and scattered deposits resulting in double-contour formation (Figure 6e). Scattered subepithelial, intramembranous, and subendothelial deposits occurred along the glomerular capillary walls. Segmental endocapillary proliferation by mononuclear cells was also present (Figure 6f). The complement alternative pathway plays an important role in the development and regulation of glomerular and tubulointerstitial disease. Specifically, this pathway contributes to the proliferative glomerulonephritis that develops in MRL/lpr mice. Hence, MRL/lpr mice deficient in either factor B8.Watanabe H. Garnier G. Circolo A. et al.Modulation of renal disease in MRL/lpr mice genetically deficient in the alternative complement pathway factor B.J Immunol. 2000; 164: 786-794Crossref PubMed Scopus (161) Google Scholar or factor D9.Elliott M.K. Jarmi T. Ruiz P. et al.Effects of complement factor D deficiency on the renal disease of MRL/lpr mice.Kidney Int. 2004; 65: 129-138Abstract Full Text Full Text PDF PubMed Scopus (86) Google Scholar showed protection against the development of glomerulonephritis in comparison to complement-sufficient MRL/lpr mice. Uncontrolled alternative pathway activation due to deficiency of factor H results in MPGN type II in humans,10.Levy M. Halbwachs-Mecarelli L. Gubler M.C. et al.H deficiency in two brothers with atypical dense intramembranous deposit disease.Kidney Int. 1986; 30: 949-956Abstract Full Text PDF PubMed Scopus (136) Google Scholar pigs,11.Hogasen K. Jansen J.H. Mollnes T.E. et al.Hereditary porcine membranoproliferative glomerulonephritis type II is caused by factor H deficiency.J Clin Invest. 1995; 95: 1054-1061Crossref PubMed Scopus (176) Google Scholar and mice.7.Pickering M.C. Cook H.T. Warren J. et al.Uncontrolled C3 activation causes membranoproliferative glomerulonephritis in mice deficient in complement factor H.Nat Genet. 2002; 31: 424-428Crossref PubMed Scopus (386) Google Scholar Activation of the alternative pathway is also required for tubulointerstitial injury to develop following ischemia–reperfusion injury as evidenced by the demonstration that mice deficient in factor B12.Thurman J.M. Ljubanovic D. Edelstein C.L. et al.Lack of a functional alternative complement pathway ameliorates ischemic acute renal failure in mice.J Immunol. 2003; 170: 1517-1523Crossref PubMed Scopus (200) Google Scholar or WT mice treated with an anti-factor B antibody13.Thurman J.M. Royer P.A. Ljubanovic D. et al.Treatment with an inhibitory monoclonal antibody to mouse factor b protects mice from induction of apoptosis and renal ischemia/reperfusion injury.J Am Soc Nephrol. 2006; 17: 707-715https://doi.org/10.1681/ASN.2005070698Crossref PubMed Scopus (95) Google Scholar show protection against tubulointerstitial damage following ischemia–reperfusion injury. Because Factor D is an essential component of the alternative pathway, we investigated the renal consequences of deficiency of this protein. Cfd−/− mutant mice develop extensive mesangial immune deposits that stain predominantly for C3 and IgM. The deposits are present in the mesangium and there is mild mesangial expansion with mild mesangial proliferative features. Our data show that the increase in mesangial cellularity is due to mesangial cell proliferation rather than infiltration by extrinsic cells. Immune deposits are absent along glomerular capillary walls. These findings are in sharp contrast to those in mice deficient in factor H, where immune deposits develop along glomerular capillary walls in association with significant proliferative changes as reported previously.7.Pickering M.C. Cook H.T. Warren J. et al.Uncontrolled C3 activation causes membranoproliferative glomerulonephritis in mice deficient in complement factor H.Nat Genet. 2002; 31: 424-428Crossref PubMed Scopus (386) Google Scholar,14.Turnberg D. Cook T.H. Complement and glomerulonephritis: new insights.Curr Opin Nephrol Hypertens. 2005; 14: 223-228Crossref PubMed Scopus (44) Google Scholar Several components of complement are expressed in the kidney. Cfd and properdin are expressed at relatively high levels in glomeruli, whereas factor B expression is significantly greater in the medulla.5.Song D. Zhou W. Sheerin S. Sacks S. Compartmental localization of complement component transcripts in the normal human kidney.Nephron. 1998; 78: 15-22Crossref PubMed Scopus (32) Google Scholar,15.Welch T. Beischel L. Frenzke M. Witte D. Regulated expression of complement factor B in the human kidney.Kidney Int. 1996; 50: 521-525Abstract Full Text PDF PubMed Scopus (20) Google Scholar Components C2, C3, and C4 and factor H are expressed predominantly in cortical tubule-rich fractions. These differences raise the possibility that there are differing roles for the alternative pathway (AP) and classical complement pathways in different regions within the kidney,5.Song D. Zhou W. Sheerin S. Sacks S. Compartmental localization of complement component transcripts in the normal human kidney.Nephron. 1998; 78: 15-22Crossref PubMed Scopus (32) Google Scholar and also suggest that factor D may play a role in local glomerular activation of the AP. Consistent with this hypothesis, we found that factor D deficiency is associated with localized alteration of the AP of complement resulting in glomerular complement and immune deposition. The renal lesions in Cfd−/− mice are restricted to the mesangium and are reminiscent of lesions seen in systemic lupus (mesangial lupus nephritis, WHO Class II) raising the possibility that there maybe an autoantibody response in Cfd−/− mice. Also, antinuclear autoimmunity has previously been noted in (129/Sv × C57BL/6) mice. Bygrave et al.16.Bygrave A.E. Rose K.L. Cortes-Hernandez J. et al.Spontaneous autoimmunity in 129 and C57BL/6 mice—implications for autoimmunity described in gene-targeted mice.PLoS Biol. 2004; 2: e243Crossref PubMed Scopus (165) Google Scholar have shown that a 129/Sv interval on chromosome 1 in otherwise completely C57BL/6 genome will drive autoimmunity, that is antinuclear antibodies. Similarly, spontaneous nephritis can be driven by a 129Sv interval on chromosome 7 on the C57BL/6 genome.17.Heidari Y. Bygrave A.E. Rigby R.J. et al.Identification of chromosome intervals from 129 and C57BL/6 mouse strains linked to the development of systemic lupus erythematosus.Genes Immun. 2006; 7: 592-599Crossref PubMed Scopus (36) Google Scholar In our study, the Cfd−/− mice have been backcrossed onto C57BL/6 but still carry the 129/Sv interval around the targeted locus. It is unlikely that the mesangial deposits present in Cfd−/− mice are due to autoimmunity, because the factor D gene is on chromosome 10, which to our knowledge is not linked with autoimmunity. It should be kept in mind that mesangial immune deposits are usually slow to develop and do not elicit the same inflammatory response as subendothelial deposits. Although mesangial deposits were noted at 4 months in Cfd−/− mutant mice (data not shown), we followed Cfd−/− mutant mice for 8 months for better identification and confirmation of the mesangial deposits. In particular, Cfd−/− mice may serve as a model for mesangial immune complex glomerulonephritis. There are only few models of mesangial immune complex disease in the literature, and these usually involve treatment of mice with various agents such as mercury,18.Martinsson K. Hultman P. The role of Fc-receptors in murine mercury-induced systemic autoimmunity.Clin Exp Immunol. 2006; 144: 309-318Crossref PubMed Scopus (17) Google Scholar mycotoxin deoxynivalenol,19.Jia Q. Shi Y. Bennink M.B. Pestka J.J. Docosahexaenoic acid and eicosapentaenoic acid, but not {alpha}-linolen.J Nutr. 2004; 134: 1353-1361PubMed Google Scholar or even intravenous injection of immune complexes.20.Leung J.C.K. Chan L.Y.Y. Tsang A.W.L. et al.Nephrol Dial Transplant. 2004; 19: 1976-1985Crossref PubMed Scopus (32) Google Scholar Our findings are somewhat surprising because deficiency of factor D would be predicted to inhibit complement activation due to decreased C3 convertase generation thereby preventing immune deposition in glomeruli. However, restriction of the immune deposits to the mesangium suggests that there is a localized protective effect of factor D. We also found that Cfd+/− mice develop some degree of mesangial deposits, although the findings are not consistent (data not shown). Possible explanations for our findings include firstly that the absence of factor D prevents intact C3 cleavage following alternative pathway activation or tick-over activation (autoactivation), and hence intact C3 accumulates at sites of synthesis, for example mesangial cells and hepatocytes. Consistent with this possibility, we found increased C3 both in the kidney and liver. Over time, this increase is harmful in the kidney, resulting in mesangial cell injury and a proliferative response (increase in mesangial matrix and increase in mesangial cells). It is also possible that lack of iC3b production in Cfd−/− mice leads to negative feedback and increased C3 synthesis. Higher circulating C3 levels6.Xu Y. Ma M. Ippolito G.C. et al.Complement activation in factor D-deficient mice.Proc Natl Acad Sci. 2001; 98: 14577-14582Crossref PubMed Scopus (127) Google Scholar presumably can cause greater C3-mediated injury to cells following immunoglobulin binding. Defining the nature of the C3 (intact C3, C3b, iC3b, and C3c) in the mesangium will shed more information on the mechanism of factor D deficiency-induced mesangial C3 and immune deposits. Second, the alternative pathway is known to be necessary for the solubilization of preformed large immune complexes.21.Schifferli J. Morris S. Dash A. Peters D. Complement-mediated solubilization in patients with systemic lupus erythematosus, nephritis or vasculitis.Clin Exp Immunol. 1981; 46: 557-564PubMed Google Scholar This process presumably occurs after tissue deposition of immune complexes and results in covalent binding of C3 to the immune complex lattice, which leads to their solubilization. In the absence of factor D, the alternative pathway is unable to solubilize mesangial immune complexes. Factor D deficiency has been reported in one family with a history of Neisseria meningitides infection.22.Biesma D.H. Hannema A.J. van Velzen-Blad H. et al.A family with complement factor D deficiency.J Clin Invest. 2001; 108: 233-240Crossref PubMed Scopus (79) Google Scholar One affected family member also developed renal failure, although this was likely due to sepsis and acute tubular necrosis. Renal biopsy was not done. In summary, our analysis demonstrates that murine factor D deficiency provides a novel model of mesangial immune complex glomerulonephritis. Further studies are now required to understand the mechanism through which factor D deficiency results in mesangial disease. For this study, 8-month-old C57BL/6 factor D-deficient mice (Cfd−/−; n=5, three males and two females), C57BL/6 factor H-deficient mice (Cfh−/−; n=6, three males and three females), and C57BL/6 WT mice (WT; n=5, two males and three females) were used. All procedures were approved by the University Animal Care and Use Committee (UACUC) of the University of Iowa. WT, Cfd−/−, and Cfh−/− mice were obtained6.Xu Y. Ma M. Ippolito G.C. et al.Complement activation in factor D-deficient mice.Proc Natl Acad Sci. 2001; 98: 14577-14582Crossref PubMed Scopus (127) Google Scholar,7.Pickering M.C. Cook H.T. Warren J. et al.Uncontrolled C3 activation causes membranoproliferative glomerulonephritis in mice deficient in complement factor H.Nat Genet. 2002; 31: 424-428Crossref PubMed Scopus (386) Google Scholar and backcrossed on to a C57BL/6 background for several generations. Identification of mouse genotypes were done using PCR as described.6.Xu Y. Ma M. Ippolito G.C. et al.Complement activation in factor D-deficient mice.Proc Natl Acad Sci. 2001; 98: 14577-14582Crossref PubMed Scopus (127) Google Scholar,7.Pickering M.C. Cook H.T. Warren J. et al.Uncontrolled C3 activation causes membranoproliferative glomerulonephritis in mice deficient in complement factor H.Nat Genet. 2002; 31: 424-428Crossref PubMed Scopus (386) Google Scholar DNA was extracted from mouse tail snips using the DNeasy Tissue Kit (Qiagen, Valencia, CA, USA). PCR amplification at 55°C annealing time for 35 cycles resulted in either a 500 bp WT sequence or 400 bp insert sequence which corresponds to the disrupted sequence of exon 3 of factor H gene or factor D gene (Figure 1). Expression patterns of factor H, factor D, and C3 were determined in mouse kidney and liver tissues using reverse transcriptase (RT)–PCR. Most complement factors are primarily produced in the liver except for factor D, which is produced in fat cells. RNA was extracted from mouse tissues using the Protein and RNA Isolation System (PARIS; Ambion). RT–PCR was carried out using the Advantage RT-for-PCR (BD Biosciences, Foster City, CA, USA). Specific primers flanking the targeted deletion for factor H and factor D genes were used for PCR (factor H forward: CTCCTGGTCAGAACAACTATATCCAG; factor H reverse: AGTTCTGTCACAGGTAGACACTTCAC; factor D forward: CTGACAGCCTTGAGGACGACCTCATTC; factor D reverse: CACACATCATGTTAATGGTGACTACC). Primers for C3 were also generated (forward: CACTCACGTAGTGACATGGTAGAGG; reverse: CTAGCTCCAATCAGGGTGTAGTAAGC) GAPDH expression was used as a control. RT–PCR studies were performed three times. The data shown in Figure 2 are representative. Mice were housed in individual metabolic cages before being killed to collect 24-h urine samples. At the time of killing, blood was obtained to assess renal function by measuring plasma creatinine and albumin; urinary creatinine and albumin were also measured (QuantiChrom™ Creatinine and BCP Albumin Assay Kits, BioAssays Systems, Hayward, CA, USA). Creatinine clearance was used as an index of glomerular filtration rate and was calculated by the following formula: Creatinine clearance=(urinary creatinine × urine volume)/serum creatinine. Mice were killed with pentobarbital (150 mg/kg intraperitoneally) followed by cervical dislocation. Kidneys were preserved in 10% formaldehyde for light microscopy, with the exception of the poles of the kidneys, which were preserved in 2.5% glutraldehyde for ultrastructural analyses by electron microscopy. A portion of each kidney was also snap-frozen for immunofluorescence studies. Formalin-fixed tissue was embedded in paraffin and 4 μM sections were stained with hematoxylin and eosin, periodic acid Schiff, and silver stain for histological analysis. The total number of nucleated cells in a glomerulus was counted. This included endothelial cells, mesangial cells, visceral epithelial cells, and any infiltrating cells. For each mouse kidney section, at least 20 glomeruli were available for counting. The total number of cells counted was divided by the total number of glomeruli counted to yield the average number of cells/glomerulus. Immunochemical staining on paraffin sections using antibodies to mouse smooth muscle actin (Dako Corporation, Carpinteria, CA, USA) to identify mesangial cells, and F4/80 (Sigma, St Louis, MO, USA) to identify monocytes/macrophages, was performed by standard staining techniques. Renal tissue was snap-frozen and embedded in Tissue Tek OCT (Miles Inc, Elkhart, IN, USA). Sections of 4 μM were cut and direct immunofluorescence studies were done using fluorescein isothiocynate-conjugated sheep antibody against IgM, IgG, IgA, and C3c (Sigma; and Binding Site, CA, USA). Semithin (1 μM) sections were stained with toluidine blue and examined. Ultrathin sections were stained with uranyl acetate and lead citrate, and examined using Jeol 100S electron microscope.