Complement-Mediated Dysfunction of Glomerular Filtration Barrier Accelerates Progressive Renal Injury

BASIC RESEARCH www.jasn.org

Mauro Abbate,* Carla Zoja,* Daniela Corna,* Daniela Rottoli,* Cristina Zanchi,* Nadia Azzollini,* Susanna Tomasoni,* Silvia Berlingeri,* Marina Noris,* Marina Morigi,* and Giuseppe Remuzzi*†

*Mario Negri Institute for Pharmacological Research and †Unit of Nephrology and Dialysis, Azienda Ospedaliera Ospedali Riuniti di Bergamo, Bergamo, Italy

ABSTRACT Intrarenal complement activation leads to chronic tubulointerstitial injury in animal models of proteinuric nephropathies, making this process a potential target for therapy. This study investigated whether a C3-mediated pathway promotes renal injury in the protein overload model and whether the abnormal exposure of proximal tubular cells to filtered complement could trigger the resulting inflammatory response. Mice with C3 deficiency were protected to a significant degree against the protein overload–induced interstitial inflammatory response and tissue damage, and they had less severe podocyte injury and less proteinuria. When the same injury was induced in wild-type (WT) mice, antiproteinuric treatment with the angiotensin-converting enzyme inhibitor lisinopril reduced the amount of plasma protein filtered, decreased the accumulation of C3 by proximal tubular cells, and protected against interstitial inflammation and damage. For determination of the injurious role of plasma-derived C3, as opposed to tubular cell–derived C3, C3-deficient kidneys were transplanted into WT mice. Protein overload led to the development of glomerular injury, accumulation of C3 in podocytes and proximal tubules, and tubulointerstitial changes. Conversely, when WT kidneys were transplanted into C3-deficient mice, protein overload led to a more mild disease and abnormal C3 deposition was not observed. These data suggest that the presence of C3 increases the glomerular filtration barrier’s susceptibility to injury, ultrafiltered C3 contributes more to tubulointerstitial damage induced by protein overload than locally synthesized C3, and local C3 synthesis is irrelevant to the development of proteinuria. It is speculated that therapies targeting complement combined with interventions to minimize proteinuria would more effectively prevent the progression of renal disease.

J Am Soc Nephrol 19: 1158–1167, 2008. doi: 10.1681/ASN.2007060686

Chronic kidney diseases (CKD) are a major global this injury, the dysfunction of the physiologic fil- and increasing health epidemic, with expected cu- tering barrier against the loss of plasma proteins mulative costs of dialysis and kidney transplanta- into the urine may act as an independent caus- tion exceeding $1 trillion in the next decade.1 Epi- ative factor.7 Abnormally ultrafiltered proteins demiology shows tight association between CKD alter gene expression in proximal tubular cells, and cardiovascular disease, predisposition of the patients to cardiovascular events, and albuminuria Received June 20, 2007. Accepted January 11, 2008. as an independent risk factor for complications and 2–4 Published online ahead of print. Publication date available at premature death. Experimental evidence indi- www.jasn.org. cates that once diseases of various etiology destroy a Correspondence: Dr. Mauro Abbate, Mario Negri Institute for critical amount of the renal mass, injury develops in Pharmacological Research, Via Gavazzeni 11, 24125 Bergamo, remaining nephrons as a consequence of maladap- Italy. Phone: ϩ39-035-319888; Fax: ϩ39-035-319331; E-mail: tive increases in intraglomerular capillary pressure [email protected] and flow.5,6 In addition to playing a pivotal role in Copyright ᮊ 2008 by the American Society of Nephrology

1158 ISSN : 1046-6673/1906-1158 J Am Soc Nephrol 19: 1158–1167, 2008 www.jasn.org BASIC RESEARCH

Figure 1. Effect of C3 deficiency on protein overload nephropathy in mice. (A and B) Urinary protein excretion (A) and tubular Ϫ Ϫ damage (B) in WT mice (n ϭ 5) and C3 / mice (n ϭ 5) receiving saline and in mice that were administered daily intraperitoneal Ϫ Ϫ injections of BSA (15 mg/g) for 4 wk (WT-BSA, n ϭ 9; C3 / BSA, n ϭ 7). ЊЊP Ͻ 0.01 versus WT-saline; **P Ͻ 0.01 versus WT-BSA. (C) Periodic acid-Schiff–stained sections of renal cortex of WT and Ϫ Ϫ C3 / mice given saline or BSA. Magnification, ϫ150. Figure 2. C3-dependent interstitial inflammatory and fibrogenic ϩ response to tubular protein overload. (A and B) Interstitial F4/80 ϩ Ϫ Ϫ resulting in proinflammatory and fibrogenic phenotypic cells (A) and ␣-SMA cells (B) in renal cortex of WT and C3 / changes.8 Angiotensin-converting enzyme inhibitors mice receiving saline or BSA. Data are means Ϯ SEM. ЊP Ͻ 0.05, ЊЊP Ͻ 0.01 versus WT-saline; *P Ͻ 0.05,**P Ͻ 0.01 versus (ACEi) and angiotensin receptor blockers prevent the pro- Ϫ Ϫ WT-BSA; #P Ͻ 0.01 versus C3 / saline. HPF, high-power field gressive impairment of the barrier and delay kidney failure. ϩ (ϫ400). The photomicrographs show F4/80 cells (A) and Drugs’ renoprotective effects cannot be dissociated from ϩ ␣-SMA cells (B), as revealed by immunoperoxidase and im- their action of lowering proteinuria, although they may not munofluorescence analysis, in kidney sections of WT mice Ϫ Ϫ be achieved invariably among patients. Clarifying further given saline or BSA and of a C3 / mouse given BSA. Magni- the pathophysiologic significance of the abnormal passage of fication, ϫ250. proteins to the luminal compartment of the nephron is man- datory in the search of drug targets for renal and cardiovascular protection. Complement proteins also accumulate in renal tubules of Intrarenal complement activation is a powerful mechanism of rats with protein overload proteinuria, a model in which tubulointerstitial injury by eliciting cytotoxic and proinflamma- plasma protein toxicity is specifically studied.24 ACEi treat- tory responses.9 C3 and other complement proteins were found in ment by limiting the transglomerular passage of proteins was proximal tubules in human renal biopsy material10,11 and in an effective maneuver to reduce both high protein and C3 load kidneys of rats after extensive renal mass reduction12–14 or of tubular cells in remnant kidneys14,21; however, no studies other proteinuric models.13,15–17 Protective effects of limiting tested the role of ultrafiltered complement in tubulointerstitial complement activation were revealed using C6-deficient ani- injury. Furthermore, proximal tubular cells are able to synthe- mals16,17 or genetically modified mice overexpressing a soluble size C3 and other complement factors23,25 and, in vitro, up- C3 inhibitor18 and by pharmacologic manipulations.16,17,19,20 regulate C3 in response to serum proteins,8,9 raising the possi- Stimulatory factors underlying complement-mediated injury bility that tubular C3 synthesis may be required to activate were not yet clarified. This is an important issue because com- pathogenic pathways. plement inhibitory approaches must be targeted timely and at We used C3-deficient mice and transplant experiments to specific steps. One hypothesis is that the exposure of proximal assess whether (1) protein overload proteinuria promotes re- tubular cells to excess plasma proteins as a consequence of nal injury through a C3-mediated pathway and (2) the abnor- glomerular barrier dysfunction could play a role.13,17,21,22 With mal exposure of proximal tubular cells to ultrafiltered comple- proteinuria, a key putative factor is the excess filtration of C3 ment could be a trigger of the inflammatory response. We (molecular weight 180 kD),14 the central molecule for the analyzed effects of antiproteinuric treatment with ACEi to vali- system to exert proinflammatory potential.14,23 In rats after date further the enhanced passage of proteins across the glomer- severe reduction of the renal mass, C3 was localized to prox- ular capillary barrier as a determinant of ultrafiltered protein and imal tubular cells engaged in high protein uptake.13,14,21 complement challenge of tubular cells.

J Am Soc Nephrol 19: 1158–1167, 2008 Filtered Protein Toxicity via C3 1159 BASIC RESEARCH www.jasn.org

droplet accumulation, cytoplasmic vacu- olization, effacement of foot processes, and formation of microvilli (Figure 3B). Podocyte damage was markedly attenuated in glomeruli of C3Ϫ/Ϫ BSA mice (Fig- ure 3C). Confocal microscopy showed absence of C3 staining in the glomeruli of WT mice given saline (Figure 3D) and, in contrast, C3 fixation of WT-BSA glomeruli, with a granular pattern corresponding to intracellular droplets of podocytes in the epithelial cell areas (Figure 3E). To test whether C3 deficiency may Figure 3. Podocyte injury is associated with C3 fixation in the absence of electron-dense be protective against tubulointerstitial Ϫ Ϫ immune deposits in WT mice with BSA overload and is attenuated in C3 / BSA mice. (A changes by abolishing the tubular load of through C) Electron micrographs show glomerular capillaries of WT mice given saline (A) C3 and other plasma proteins, which acts Ϫ/Ϫ or BSA (B) and of a C3 BSA mouse (C). The WT-BSA mouse glomerulus (B) shows as a trigger of tubular cell activation and podocyte swelling and protein droplets (arrows) and foot process fusion in the absence Ϫ Ϫ proinflammatory phenotypic change, we of electron-dense deposits. Podocyte changes are less marked in C3 / BSA (C). (D and performed immunohistochemical stud- E) In contrast to WT mouse given saline (D) showing only interrupted staining of Bow- man’s capsule (BC), the detection of C3 by confocal microscopy in a WT-BSA glomerulus ies for detection of filtered plasma pro- reveals C3 fixation of WT-BSA glomeruli, with a granular pattern corresponding to teins in tubular epithelial cells. In kid- intracellular droplets of podocytes in glomerular epithelial cell areas (E, arrows). Magni- neys of WT mice, C3 had peritubular fications: ϫ5600 in A through C; ϫ1000 in D and E. distribution with no apical or intracellular staining (Figure 4A). In contrast, in RESULTS WT-BSA mice, C3 was detected on apical cell surface and within proximal tubular cells in addition to the normal peri- C3 Deficiency Protects Mice against Renal Damage tubular staining (Figure 4A). No IgG was found in tubular Ϫ Ϫ Induced by Protein Overload cells of WT (Figure 4B) or C3 / controls (data not shown). To assess the role of C3 in the pathogenesis of proteinuria- Kidneys of C3 Ϫ/ϪBSA mice showed abnormal staining of induced tubulointerstitial damage, we first characterized the IgG in proximal tubules, albeit to a lesser degree with re- renal disease phenotype of protein overload proteinuria in C3 spect to WT-BSA mice (Figure 4B). Kidneys of WT-BSA gene–disrupted (C3Ϫ/Ϫ) mice. Wild-type (WT) mice given mice killed at an early stage of disease (week 1) showed BSA (WT-BSA) for 4 wk had high levels of urinary protein abnormal staining for C3 and IgG in proximal tubules, be- excretion (Figure 1A). WT-BSA mice developed tubular lumi- fore the onset of tubular changes (Figure 5). These findings nal casts and dilation and focal glomerular sclerotic lesions indicate that C3 deficiency may exert protective effects on (Figure 1, B and C). C3Ϫ/Ϫ mice given BSA (C3Ϫ/ϪBSA) had the tubulointerstitium in the presence of residual dysfunc- lower degrees of proteinuria (Figure 1A) and tubular damage tion of the glomerular barrier, without fully abrogating the (Figure 1, B and C). Percentages of glomeruli showing sclerotic exposure of proximal tubular cells to ultrafiltered proteins. changes were lower in C3Ϫ/ϪBSA than in WT-BSA mice (0.3 Ϯ Ϫ Ϫ versus 5 Ϯ 1; P Ͻ 0.01). Control WT or C3 / mice given Effects of Transplantation of C3؊/؊ Kidneys into WT 0.2 saline showed no changes. Structural changes in WT-BSA mice mice or WT Kidneys into C3؊/؊ Mice on Renal Injury were associated with accumulation of F4/80-positive cells (Fig- of Protein Overload Proteinuria ure 2A) and ␣-smooth muscle actin (␣-SMA)-positive cells We tested the role of plasma-derived C3 by transplantation (Figure 2B) into the interstitium. As compared with WT-BSA experiments in which congenic C3Ϫ/Ϫ kidneys were trans- mice, the interstitium of C3Ϫ/Ϫ BSA mice contained signifi- planted into WT recipients before exposure to protein over- cantly lower numbers of F4/80-positive cells and ␣-SMA–pos- load proteinuria, and vice versa. WT mice given WT kidneys itive cells. Given the protection exerted by C3 deficiency developed protein overload proteinuria in response to BSA against proteinuria and glomerular sclerosis, we performed injections (Table 1). Proteinuria (Table 1) and tubular injury studies to assess the effects of C3 deficiency on the structural (score 1.00 Ϯ 0.00 versus 0.96 Ϯ 0.04) of WT mice that received integrity of the glomerular barrier in this model. Neither cir- a transplant of C3-deficient kidneys were comparable to those culating anti-BSA antibodies (data not shown) nor electron- found in the WT mice that received a transplant of WT kidneys dense immune deposits or infiltrating immune-type cells were in this model. The percentage of glomeruli showing sclerotic detectable in WT-BSA mice (Figure 3), thus excluding a typical lesion (mean percentage 5.6% [range 3 to 7%] and 6% [range form of immunologically mediated glomerulonephritis. WT- 5 to 7%]) was comparable in both groups, which also showed BSA mice developed podocyte damage consisting of protein marked podocyte damage and glomerular C3 fixation (Figure

1160 Journal of the American Society of Nephrology J Am Soc Nephrol 19: 1158–1167, 2008 www.jasn.org BASIC RESEARCH

Figure 5. C3 accumulation and protein overload of proximal tubules in early stage of protein overload nephropathy (week 1). Sections of renal cortex of mice that received daily intraperitoneal injections of BSA (A, C, and E) or saline (B, D, and F), stained for C3 (top) and IgG (middle) or with periodic acid-Schiff-stain (bot- tom). Magnification, ϫ250.

Figure 4. (A and B) Proximal tubular staining for C3 (A) and IgG Lisinopril Limits C3 Accumulation and C3 mRNA Ϫ Ϫ (B) in renal cortex of WT mice or C3 / mice receiving saline and Upregulation in Proximal Tubules and of mice that were administered daily intraperitoneal injections of Tubulointerstitial Changes in Mice with Protein Ϫ Ϫ BSA (WT-BSA and C3 / BSA). Data are means Ϯ SEM. ЊЊP Ͻ 0.01 Overload Proteinuria Ϫ Ϫ versus WT-saline; #P Ͻ 0.01 versus C3 / saline (peripheral C3 ACEi was previously found to limit both proteinuria and C3 was assigned score 0; see the Concise Methods section for de- deposition in proximal tubules in rats with progressive tubu- tails). Photomicrographs show C3 (A) and IgG staining (B). Mag- lointerstitial injury.14,21 To verify further excess protein load ϫ nification, 250. on proximal tubular cells as a stimulus for the activation of complement-mediated injury, we investigated whether ACEi treatment by reducing the high protein load could lead to less 6). Abnormal staining for C3 was found within the cells and on C3 accumulation in proximal tubules and less tubulointersti- apical surface in proximal tubules in transplanted kidneys tial damage. Lisinopril treatment significantly reduced both (Figure 7), to similar extent in both groups (Table 1). These proteinuria and the proximal tubular cell staining of C3 and changes were associated with protein overload of proximal tu- IgG as compared with untreated mice with BSA overload (Ta- bules, as reflected by evidence of IgG accumulation at this site ble 2). Lisinopril reduced tubular damage score and numbers (data not shown). F4/80ϩ cells into the interstitium were lower of F4/80ϩ cells in the renal interstitium (Table 2). but not to a significant extent in transplanted C3Ϫ/Ϫ kidneys as To establish whether protein overload may increase C3 syn- compared with WT kidneys (Table 1, Figure 7). C3Ϫ/Ϫ mice thesis by tubular cells, we first evaluated by real-time reverse that received a transplant of WT kidneys and were exposed to transcriptase–PCR (RT-PCR) the effects of BSA overload on BSA overload had less severe proteinuria (Table 1) and tubular C3 mRNA in whole kidneys. BSA overload caused a 14-fold injury (score 0.16 Ϯ 0.16; P Ͻ 0.05 versus WT/WT BSA mice upregulation of C3 mRNA (P Ͻ 0.05 versus saline-injected and KO/WT BSA mice), a lower percentage of glomeruli show- mice; Figure 8A). In situ hybridization analysis showed focal ing sclerosis (0.33%; range 0 to 1%; P Ͻ 0.05 versus WT/WT C3 mRNA staining in tubular epithelial cells of kidney sections BSA mice and KO/WT BSA mice), and reduced interstitial of WT control mice that were administered an injection of F4/80ϩ cell accumulation (Table 1, Figure 7). The protective saline (Figure 8B). A strong increase in C3 mRNA was detected effects of systemic C3 deficiency were associated with less in tubular epithelial cells of both proximal tubules and distal prominent podocyte injury (Figure 6), in the absence of abnor- segments in WT-BSA mice (Figure 8B). No change was found mal staining of C3 both in the glomeruli (Figure 6) and at in kidneys of WT-BSA mice killed at early stage (1 wk; Figure tubular level (Table 1, Figure 7). 8B). Sense probe produced no or trace nonspecific staining.

J Am Soc Nephrol 19: 1158–1167, 2008 Filtered Protein Toxicity via C3 1161 BASIC RESEARCH www.jasn.org

Ϫ Ϫ Ϫ Ϫ Table 1. Effects of C3 / kidney transplantation in WT mice or WT kidney transplantation in C3 / mice on proteinuria, ϩ accumulation of C3 in proximal tubules, and interstitial F4/80 cell infiltration of BSA overload nephropathya (Group Proteinuria (mg/d) Tubular C3 Staining (Score) F4/80؉ cells (n/HPF WT/WT ϩ BSA 27 Ϯ 3 0.75 Ϯ 0.12 91 Ϯ 14 KO/WT ϩ BSA 32 Ϯ 4 0.90 Ϯ 0.30 62 Ϯ 4 b WT/KO ϩ BSA 13 Ϯ 4b 0.00 Ϯ 0.00 31 Ϯ 6 aData are means Ϯ SEM. Tubular C3 staining refers to abnormal intratubular C3 accumulation (see the Concise Methods section). WT mice received a Ϫ Ϫ Ϫ Ϫ transplant of WT kidneys (WT/WT; n ϭ 5) or C3 / kidneys (KO/WT; n ϭ 3), and C3 / mice received a transplant of WT kidneys (WT/KO; n ϭ 3) before BSA injections. HPF, high-power field (ϫ400). bP Ͻ 0.05 versus WT/WT ϩ BSA and KO/WT ϩ BSA.

C3Ϫ/Ϫ kidneys had no signal (data not shown). Lisinopril major novel finding is that despite no evidence of antibody-me- caused a significant reduction of whole kidney C3 mRNA levels diated nephritis, the degrees of proteinuria and glomerular dam- as compared with BSA mice given vehicle (Figure 8A). In con- age were low in C3Ϫ/Ϫ mice, suggesting that the injury to the trast to the strong signal for C3 mRNA that was focally detect- glomerular barrier was exacerbated by complement and that the able by in situ hybridization in this group, focal perinuclear C3 related stimulation of proximal tubular cells was reduced in its mRNA staining was found in proximal tubular cells in kidneys absence. Protection by systemic C3 deficiency was exerted even on of lisinopril-treated BSA mice (Figure 8B). transplanted kidneys with intact local complement system. Col- lectively, our data document that complement is recruited locally to augment both podocyte injury and the interstitial inflamma- DISCUSSION tory response and that protein overload is a stimulatory factor. Candidate effector molecules of renal epithelial cell injury have The search for mechanisms of progressive renal injury has crucial been recognized both in experimental models19,25 and in vitro.26 importance in renal and cardiovascular medicine. Protein over- Alternative pathway complement activation through C3 conver- load proteinuria is a unique ad hoc model whereby the role of tase leads to C5b-9 (MAC) binding on tubular cell surface12 and toxicity of filtered proteins in tubulointerstitial damage can be causes tubular cell dysfunction via reactive oxygen species and Ϫ Ϫ investigated in mice. Findings that protein-overloaded C3 / cytokines.26 A monocyte-activating amidated form of C3,12 mice were significantly protected from tubular injury and inter- C3a,27,28 and C529 can also be pathogenic. stitial accumulation of macrophages and myofibroblasts indicate Glomerular permselective dysfunction in the protein overload that C3 plays a pivotal role in processes by which the abnormal model leads to urinary loss of endogenous plasma proteins.30 passage of plasma proteins across the glomerular barrier into the Consistently, the attenuation of podocyte injury accounted for luminal compartment of the nephron damages the kidney. One reduction of proteinuria by C3 deficiency. Previous investigation showed that lisinopril treatment by pre- serving the integrity of the glomerular bar- rier31,32 limited the exposure of the tubule to plasma proteins and attenuated both C3 accumulation in proximal tubular cells and renal injury in rat remnant kidneys.14,21 Similar effects were achieved here by lisinopril manipulation, further indicat- ing that protein overload acts as a stimulus for complement accumulation in tubular cells and the associated interstitial inflammatory reaction. In light of less tubulointerstitial damage found in C3Ϫ/Ϫ mice, reducing the accumulation of complement proteins in proximal tubular cells, or at least the apical exposure of the cells to plas- Figure 6. Electron micrographs of podocytes and immunofluorescence staining for C3 ma-derived C3, is in all likelihood a com- in glomeruli of transplanted kidneys in BSA overload. WT mice received a transplant of Ϫ/Ϫ Ϫ/Ϫ ponent of the drug’s protective action either a WT kidney (left) or a C3 kidney (middle), and the C3 mouse received a against tubulointerstitial injury. Evidence transplant of a WT kidney (right), followed by BSA injections. (A through F) In both the Ϫ Ϫ that complement acts downstream of pro- transplanted WT kidney (A) and C3 / kidney (B) in WT hosts, podocytes show prominent protein droplets and fusion of foot processes and abnormal C3 fixation (D and E). In tein load to enhance tubulointerstitial in- contrast, the glomerulus of the transplanted WT kidney in C3-deficient host shows less jury also explains less susceptibility of C6- severe podocyte injury (C) and no abnormal C3 fixation (F).Magnifications: ϫ5600 in A deficient rats to tubulointerstitial damage through C; ϫ800 in D through F. despite proteinuria in aminonucleoside of

1162 Journal of the American Society of Nephrology J Am Soc Nephrol 19: 1158–1167, 2008 www.jasn.org BASIC RESEARCH

into C3Ϫ/Ϫ mice in the presence of residual proteinuria; however, the abundance of C3 reaching the ultrafiltrate seems to overwhelm locally synthesized C3. Con- sistently, C3 overload of proximal tubular cells was present since a stage of disease in which tubular C3 mRNA was not upregulated. Collectively, our data show that limiting glomerular barrier dysfunction attenuates the excess protein load responsible for abnormal accumulation of C3 in the renal tu- Figure 7. Effects of selectively abrogating intrarenal synthesis of C3 versus circulating bule. The finding of significant reduction ϩ Ϫ/Ϫ C3 on the accumulation of C3 in proximal tubules and interstitial F4/80 cells of BSA of proteinuria in C3 mice with BSA overload nephropathy. WT mice received a transplant of either a WT kidney (left) or a overload makes it difficult to evaluate to Ϫ Ϫ Ϫ Ϫ C3 / kidney (middle), and a C3 / mouse received a transplant of a WT kidney (right), which extent the reduced tubular load with followed by BSA overload. Staining of kidney sections for C3 (A through C) or F4/80 (D other plasma proteins and the lack of fil- through F) shows abnormal intratubular C3 deposition (A and B) and interstitial macro- tered C3 itself, respectively, concurred to phage infiltration (D and E) in response to BSA overload in WT mice. The lack of abnormal ϩ protect against tubulointerstitial injury. deposition of plasma-derived C3 (C) is associated with less severe F4/80 cell infiltration Ϫ/Ϫ ϫ Both in C3 mice and in ACEi-treated (F) in transplanted WT kidney. Magnification, 250. mice, residual accumulation of plasma proteins in proximal tubules and tubuloin- puromycin16 or renal mass reduction.17 Instead, no apical com- terstitial changes reflects the contribution of complement-inde- plement deposition in tubules occurred and the lack of C6 was not pendent mechanisms of injury in the presence of excess protein in protective in nonproteinuric models,22 indirectly suggesting the the ultrafiltrate. They include proximal tubule cell synthesis of importance of apical exposure to complement to enhance injury mediators (monocyte chemoattractant protein-1, RANTES, frac- in proteinuric nephropathy. talkine, endothelin-1, and TGF-␤1) and/or other stimulatory ef- C3 (180 kD) is mostly circulating and of hepatic source; fects by protein-bound molecules.7 In humans, data using laser however, a role in renal injury has been suggested for the ca- capture microdissection and microarray analysis revealed com- pacity of kidney cells to synthesize C3, C4, C5, and factor B.25,33 plex gene regulatory responses of proximal tubular cells in pro- C3 of proximal tubule cell origin is a crucial mediator of injury teinuric nephropathies.38 Cultured proximal tubular cells via the in experimental acute allograft rejection and after renal isch- megalin receptor for protein uptake can undergo activation by emia and reperfusion.23,34 Its expression was enhanced in both regulated intramembrane proteolysis39 and cellular damage.40 human and experimental proteinuric nephropathies.35 More- Whether these responses may be linked to and possibly modify over, exposing cultured proximal tubular cells to serum pro- effects of complement pathway(s) is not known. It is noteworthy teins at the apical surface enhanced C3 mRNA expression36 that, in vitro, excess protein exposure enhanced complement ac- and secretion.36,37 Conversely, our transplantation studies tivation via alternative pathway in human proximal tubular done to abrogate C3 synthesis by the kidney show that C3 cells.41 accumulates in renal tubules to promote tubulointerstitial in- A key observation of this study is that preventing the accu- jury, whereas the latter is less severe in systemically C3-defi- mulation of C3 in podocytes was associated with less severe cient mice with transplanted WT kidneys. These data would podocyte damage, both in C3Ϫ/Ϫ mice with BSA overload and not exclude the pathogenic potential of tubular C3 synthesis in WT kidneys transplanted in C3Ϫ/Ϫ mice in this model, re- on more severe glomerular injury and proteinuria as found in sulting in protection against progressive glomerular injury. WT mice. Here, lisinopril prevented C3 mRNA upregulation Thus, whereas locally synthesized C3 seems to be irrelevant to in tubular cells of BSA-overloaded mice. In addition, peritu- the pathogenesis of proteinuria, the presence of circulating bular C3 staining was preserved in WT kidneys transplanted C3 acts as a susceptibility factor that negatively influences pro-

Table 2. Effects of lisinopril treatment on proteinuria, tubular damage, and C3 and IgG accumulation in proximal tubules ϩ and interstitial F4/80 cells in WT mice receiving BSAa Proteinuria Tubular Damage Tubular Staining (Score) (Group F4/80؉ Cells (n/HPF (mg/d) (Score) C3 IgG WT ϩ BSA (n ϭ 4) 36 Ϯ 6 1.10 Ϯ 0.20 1.3 Ϯ 0.1 2.1 Ϯ 0.2 49 Ϯ 3 WT ϩ BSA ϩ lisinopril (n ϭ 6) 12 Ϯ 4b 0.22 Ϯ 0.10c 0.5 Ϯ 0.1c 0.9 Ϯ 0.2c 20 Ϯ 5c aData are means Ϯ SEM. Tubular C3 staining refers to abnormal intratubular C3 accumulation (see the Concise Methods section). bP Ͻ 0.05, cP Ͻ 0.01 versus WT ϩ BSA.

J Am Soc Nephrol 19: 1158–1167, 2008 Filtered Protein Toxicity via C3 1163 BASIC RESEARCH www.jasn.org

with cardiopulmonary bypass.47 Hopefully, complement inhibitors will be useful to prevent progression of human proteinuric renal diseases, when combined with interventions to minimize the dysfunction of the glomerular barrier to proteins.

CONCISE METHODS

Animals Male C3-deficient mice (strain B6.129S4-C3tm1Crr) of C57BL/6 ge- netic background and age-matched C57BL/6 WT mice were created by heterozygote ϫ heterozygote breeder pairs from Jackson Labora- tory (Bar Harbor, ME). Genotypes were determined by PCR of DNA accordingly to the protocol C3tm1Crr from Jackson Laboratory. Animal care and treatment were in accordance with institutional guidelines in compliance with national (Decreto Legislativo n.116, Gazzetta Ufficiale suppl 40, 18 febbraio 1992, Circolare n.8, Gazzetta Ufficiale 14 luglio 1994) and international laws and policies (EEC Figure 8. Enhanced C3 mRNA expression in kidney and inhibi- Council Directive 86/609, OJL358-1, December 1987; Guide for the tion by antiproteinuric therapy in protein overload nephropathy. Care and Use of Laboratory Animals, US National Research Council, (A) Real-time RT-PCR. Data are means Ϯ SEM. ЊЊP Ͻ 0.01 versus 1996). Animals were housed at a constant room temperature, 12-h WT-saline; *P Ͻ 0.05 versus WT-BSA. (B) In situ hybridization on dark/12-h light cycle, and fed a standard diet. kidney sections of WT mice given saline, BSA, or BSA plus lisinopril daily (n ϭ 4 mice for each group). Magnification, ϫ250. Murine Model of Protein Overload Proteinuria BSA overload proteinuria was induced in WT or C3-deficient mice. teinuria and its deleterious consequences. Our data are fully Uninephrectomy was performed 5 d before starting BSA.48,49 Low consistent with a role of protein overload of glomerular epi- endotoxin BSA (Sigma A-9430, St. Louis, MO) was given 5 d weekly thelial cells in the development of the sclerotic lesion in pro- intraperitoneally (days 1 through 5) at the dosage of 15 mg/g body wt teinuric disease.42 This possibility is strengthened by findings for 4 wk. Control WT and C3-deficient mice received the same vol- that C3 deficiency attenuated podocyte damage and sclerosis ume of saline. For establishment of whether C3 deposition may ensue in Adriamycin nephropathy in mice, and, conversely, CD59 in advance of tubular changes, two groups of mice given BSA or saline deficiency exacerbated the disease.43 Because podocytes nor- were killed at 1 wk for immunohistology. In another set of experi- mally express complement inhibitory molecules such as com- ments, beginning on day 1 of BSA injection, uninephrectomized mice plement receptor 1 or its mouse analogue factor H,44 it is rea- received daily either lisinopril (Astra Zeneca, Basiglio, Milano, Italy; sonable to suggest that perturbed complement regulation may 60 mg/L in the drinking water) or no treatment up to 4 wk. Protein- represent an important feature of this process. uria was determined as described previously.48 Improved success in preventing CKD progression could come from manipulation of intrarenal complement activation, but stimulatory factors and source(s) of complement in pro- Anti-BSA Antibodies teinuric nephropathy had not been characterized so far. We At variance with other protocols,50,51 this model does not cause im- found that C3 deficiency prevents accumulation of circulating mune complex disease.24 For further exclusion of immunologic fac- C3 in glomerular cells and attenuates podocyte injury, result- tors, plasma of mice with BSA overload proteinuria (4 wk) versus ing in protection against proteinuria and glomerular sclerosis. controls (n ϭ 3 each group) was analyzed for BSA antibodies on serial Our data provide the evidence that the abnormal passage of dilutions using the double gel diffusion method of Ouchterlony.52 C3 protein across the glomerular filtering barrier is the factor underlying the accumulation of C3 together with other filtered Mouse Kidney Transplantation proteins in proximal tubules, and systemic C3 deficiency at- WT mice received a transplant of either C3Ϫ/Ϫ kidney or WT kidney, tenuates interstitial injury. Finally, protective effects of lisino- and C3Ϫ/Ϫ mice received a transplant of WT kidney.53 The donor’s pril treatment are due to its antiproteinuric action, at least left kidney was flushed with ice-cold heparinized saline and removed partly, via reduction of tubular load by C3 and/or ultrafiltered together with ureter and vessels. Recipient mice underwent left-sided proteins. C5 antibodies blocked complement activation in nephrectomy and orthotopic implantation. The ureter was inserted phase II clinical trials in patients with paroxysmal nocturnal into the bladder and pulled through.54 The right native kidney was hemoglobinuria45 or myocardial infarction.46 A soluble form removed on day 15 postoperatively, whereupon recipients became of the C3/C5 convertase inhibitor CR1 (TP10) decreased com- dependent on a functioning graft. Animals without graft function plement activation and protected vascular function in infants died within 36 h of the contralateral nephrectomy. Protein overload

1164 Journal of the American Society of Nephrology J Am Soc Nephrol 19: 1158–1167, 2008 www.jasn.org BASIC RESEARCH

proteinuria was induced starting on day 15 after removal of the native In Situ Hybridization kidney. A 301-bp fragment of C3 cDNA was cloned into transcription vector pSPT19 between SP6 and T7 promoters. Mouse C3 antisense and Renal Histology sense riboprobes were prepared and labeled by in vitro transcription Kidneys were processed as described previously.48 Tubular (atrophy, using digoxigenin RNA labeling kit (Roche Diagnostic, Monza, Italy). casts, and dilation) and interstitial changes (fibrosis and inflamma- In situ hybridization was performed as described previously.58 Sec- tion) were graded from 0 to 3ϩ (0, no changes; 1ϩ, changes affecting tions were hybridized overnight with RNA probes at final concentra- Ͻ25% of the sample; 2ϩ, changes affecting 25 to 50% of the sample; tion of 0.5 ng/␮l, in 2ϫ SSC, 10% dextran sulfate, 1ϫ Denhardt’s 3ϩ, changes affecting Ͼ50% of the sample). At least 100 glomeruli solution, 20 mM Vanadil Ribonucleoside Complex (New England were examined for each animal. Percentages of glomeruli presenting Biolabs, Frankfurt Am Main, Germany), and 0.1 M sodium phos- sclerotic lesions were recorded. Biopsies were analyzed by one pathol- phate in a moist chamber at 50°C. ogist, who was unaware of experimental groups. Statistical Analyses Immunohistology Results are expressed as means Ϯ SEM. Statistical analysis was per- Cortical samples were embedded in OCT compound and fresh-frozen formed using nonparametric Kruskal-Wallis or Mann Whitney tests in liquid nitrogen. Acetone-fixed sections were stained (1 h) with or ANOVA followed by Tukey test for multiple comparisons. Statis- ␮ FITC-conjugated sheep anti-mouse IgG (20 g/ml; Sigma) or goat tical significance level was defined as P Ͻ 0.05. anti-mouse C3 antibody (10 ␮g/ml; Cappel, Durham, NC). Proximal tubule staining was scored as follows: 0, no IgG or linear peripheral C3; 1, intracellular and/or brush border staining in few tubular profiles (Ͻ5% in each microscopic field); 2, staining affecting 5 to 50% of ACKNOWLEDGMENTS profiles per field; 3, Ͼ50% of profiles per field. Proximal tubules were identified on the basis of morphology, presence of brush border, and Part of this work received support by a Genzyme Renal Innovation topography. At least 20 randomly selected high-power microscopic Program grant. S.B. was a recipient of a fellowship “Fondazione fields (ϫ400) were analyzed per animal. Glomerular C3 localization ARMR—The Nando Peretti Foundation.” was further assessed using an inverted confocal laser microscope Part of this work was presented at the annual meeting of the Amer- (LSM 510 meta; Zeiss, Jena, Germany). ican Society of Nephrology; October 27 through November 1, 2004; Both F4/80 and ␣-SMA have been used as end-point markers for St. Louis, MO. inflammation and fibrosis in this model.55,56 F4/80 analysis was per- We thank Drs. Daniela Macconi and Simona Buelli for helpful formed as described previously.48 The primary antibody (Caltag Lab- technical advice and discussion. Manuela Passera helped to prepare oratories, Burlingame, CA), was incubated overnight (2.5 ␮g/ml, the manuscript. 4°C), followed by biotinylated goat anti-rat IgG (Vector Laborato- ries), avidin-biotin peroxidase complex, and DAB. For detection of ␣-SMA, 3-␮m frozen sections were blocked with PBS/1% BSA and DISCLOSURES incubated (1 h, room temperature) with Cy3-conjugated anti–␣- None. SMA antibody (1A4, 1:200; Sigma). At least 20 randomly selected high-power fields per section were assigned a score: 0, no staining; 1ϩ, scattered ␣-SMA–positive cells in peritubular areas; 2ϩ, moder- REFERENCES ate ␣-SMA peritubular staining; 3ϩ, marked accumulation of cells surrounding Ͼ50% of the tubules. The investigator was unaware of 1. Perico N, Codreanu I, Schieppati A, Remuzzi G: Pathophysiology of experimental groups. disease progression in proteinuric nephropathies. Kidney Int Suppl S79–S82, 2005 Electron Microscopy 2. Sarnak MJ, Levey AS, Schoolwerth AC, Coresh J, Culleton B, Hamm LL, McCullough PA, Kasiske BL, Kelepouris E, Klag MJ, Parfrey P, Cortical kidney fragments were fixed in 2.5% glutaraldehyde in 0.1 M Pfeffer M, Raij L, Spinosa DJ, Wilson PW: Kidney disease as a risk sodium cacodylate buffer and embedded in Epon resin by standard factor for development of cardiovascular disease: A statement from methods. Glomeruli showing no sclerotic lesion (n ϭ 4 glomeruli the American Heart Association Councils on Kidney in Cardiovascular from three animals per group) were selected on semithin sections and Disease, High Blood Pressure Research, Clinical Cardiology, and Ep- idemiology and Prevention. Circulation 108: 2154–2169, 2003 further processed for analysis. 3. Hostetter TH: Chronic kidney disease predicts cardiovascular disease. N Engl J Med 351: 1344–1346, 2004 Quantitative Real-Time RT-PCR 4. Irie F, Iso H, Sairenchi T, Fukasawa N, Yamagishi K, Ikehara S, Kanashiki RT-PCR was performed as described previously.57 The following M, Saito Y, Ota H, Nose T: The relationships of proteinuria, serum primers were used: Murine C3 (300 nM) forward 5Ј-AAGCCAAG- creatinine, glomerular filtration rate with cardiovascular disease mortality in Japanese general population. Kidney Int 69: 1264–1271, 2006 GACTGCAGACTGA-3Ј and reverse 5Ј-GGCTTGGTGCACTCAA- 5. Brenner BM, Meyer TW, Hostetter TH: Dietary protein intake and the Ј Ј GATCT-3 and 18S (50 nM) forward 5 -ACGGCTACCACATC- progressive nature of kidney disease: The role of hemodynamically CAAGGA-3Ј and reverse 5Ј- CGGGAGTGGGTAATTTGCG-3Ј. mediated glomerular injury in the pathogenesis of progressive glo-

J Am Soc Nephrol 19: 1158–1167, 2008 Filtered Protein Toxicity via C3 1165 BASIC RESEARCH www.jasn.org

merular sclerosis in aging, renal ablation, and intrinsic renal disease. 26. Biancone L, David S, Della Pietra V, Montrucchio G, Cambi V, Camussi N Engl J Med 307: 652–659, 1982 G: Alternative pathway activation of complement by cultured human 6. Anderson S, Meyer TW, Rennke HG, Brenner BM: Control of glomer- proximal tubular epithelial cells. Kidney Int 45: 451–460, 1994 ular hypertension limits glomerular injury in rats with reduced renal 27. Peake PW, O’Grady S, Pussell BA, Charlesworth JA: C3a is made by mass. J Clin Invest 76: 612–619, 1985 proximal tubular HK-2 cells and activates them via the C3a receptor. 7. Remuzzi G: Nephropathic nature of proteinuria. Curr Opin Nephrol Kidney Int 56: 1729–1736, 1999 Hypertens 8: 655–663, 1999 28. Braun MC, Reins RY, Li TB, Hollmann TJ, Dutta R, Rick WA, Teng BB, 8. Abbate M, Zoja C, Remuzzi G: How does proteinuria cause progres- Ke B: Renal expression of the C3a receptor and functional responses sive renal damage? J Am Soc Nephrol 17: 2974–2984, 2006 of primary human proximal tubular epithelial cells. J Immunol 173: 9. Quigg RJ: Complement and the kidney. J Immunol 171: 3319–3324, 4190–4196, 2004 2003 29. Boor P, Konieczny A, Villa L, Schult AL, Bucher E, Rong S, Kunter U, 10. Camussi G, Stratta P, Mazzucco G, Gaido M, Tetta C, Castello R, van Roeyen CR, Polakowski T, Hawlisch H, Hillebrandt S, Lammert F, Rotunno M, Vercellone A: In vivo localization of C3 on the brush Eitner F, Floege J, Ostendorf T: Complement C5 mediates experi- border of proximal tubules of kidneys from nephrotic patients. Clin mental tubulointerstitial fibrosis. J Am Soc Nephrol 18: 1508–1515, Nephrol 23: 134–141, 1985 2007 11. Hsu SI-H, Couser WG: Chronic progression of tubulointerstitial dam- 30. Weening JJ, Van Guldener C, Daha MR, Klar N, Van Der Wal A, Prins age in proteinuric renal disease is mediated by complement activa- FA: The pathophysiology of protein-overload proteinuria. Am J Pathol tion: A therapeutic role for complement inhibitors? J Am Soc Nephrol 129: 64–73, 1987 14: S186–S191, 2003 31. Macconi D, Ghilardi M, Bonassi ME, Mohamed EI, Abbate M, Colombi 12. Nath KA, Hostetter MK, Hostetter TH: Pathophysiology of chronic tubu- F, Remuzzi G, Remuzzi A: Effect of angiotensin-converting enzyme lo-interstitial disease in rats: Interactions of dietary acid load, ammonia, inhibition on glomerular basement membrane permeability and dis- and complement component C3. J Clin Invest 76: 667–675, 1985 tribution of zonula occludens-1 in MWF rats. J Am Soc Nephrol 11: 13. Abbate M, Zoja C, Corna D, Capitanio M, Bertani T, Remuzzi G: In 477–489, 2000 progressive nephropathies, overload of tubular cells with filtered pro- 32. Macconi D, Abbate M, Morigi M, Angioletti S, Mister M, Buelli S, teins translates glomerular permeability dysfunction into cellular sig- Bonomelli M, Mundel P, Endlich K, Remuzzi A, Remuzzi G: Permselec- nals of interstitial inflammation. J Am Soc Nephrol 9: 1213–1224, 1998 tive dysfunction of podocyte-podocyte contact upon angiotensin II 14. Abbate M, Zoja C, Rottoli D, Corna D, Perico N, Bertani T, Remuzzi G: unravels the molecular target for renoprotective intervention. Am J Antiproteinuric therapy while preventing the abnormal protein traffic Pathol 168: 1073–1085, 2006 in proximal tubule abrogates protein and complement-dependent 33. Welch TR, Blystone LW: C3 is central to the interstitial component of interstitial inflammation in experimental renal disease. JAmSoc experimental immune complex glomerulonephritis. Clin Immunol 115: Nephrol 10: 804–813, 1999 80–84, 2005 15. Nomura A, Morita Y, Maruyama S, Hotta N, Nadai M, Wang L, Hase- 34. Farrar CA, Zhou W, Lin T, Sacks SH: Local extravascular pool of C3 is gawa T, Matsuo S: Role of complement in acute tubulointerstitial injury of a determinant of postischemic acute renal failure. FASEB J 20: 217– rats with aminonucleoside nephrosis. Am J Pathol 151: 539–547, 1997 226, 2006 16. Nangaku M, Pippin J, Couser WG: Complement membrane attack 35. Tang S, Lai KN, Sacks SH: Role of complement in tubulointerstitial complex (C5b-9) mediates interstitial disease in experimental ne- injury from proteinuria. Kidney Blood Press Res 25: 120–126, 2002 phrotic syndrome. J Am Soc Nephrol 10: 2323–2331, 1999 36. Tang S, Neng Lai K, Mao Chan T, Lan HY, Ho SK, Sacks SH: Transferrin 17. Nangaku M, Pippin J, Couser WG: C6 mediates chronic progression of but not albumin mediates stimulation of complement C3 biosynthesis tubulointerstitial damage in rats with remnant kidneys. JAmSoc in human proximal tubular epithelial cells. Am J Kidney Dis 37: 94– Nephrol 13: 928–936, 2002 103, 2001 18. Bao L, Haas M, Boackle SA, Kraus DM, Cunningham PN, Park P, 37. Tang S, Sheerin NS, Zhou W, Brown Z, Sacks SH: Apical proteins Alexander JJ, Anderson RK, Culhane K, Holers VM, Quigg RJ: Trans- stimulate complement synthesis by cultured human proximal tubular genic expression of a soluble complement inhibitor protects against epithelial cells. J Am Soc Nephrol 10: 69–76, 1999 renal disease and promotes survival in MRL/lpr mice. J Immunol 168: 38. Rudnicki M, Eder S, Perco P, Enrich J, Scheiber K, Koppelstatter C, 3601–3607, 2002 Schratzberger G, Mayer B, Oberbauer R, Meyer TW, Mayer G: Gene 19. He C, Imai M, Song H, Quigg RJ, Tomlinson S: Complement inhibitors expression profiles of human proximal tubular epithelial cells in pro- targeted to the proximal tubule prevent injury in experimental ne- teinuric nephropathies. Kidney Int 71: 325–335, 2007 phrotic syndrome and demonstrate a key role for C5b-9. J Immunol 39. Zou Z, Chung B, Nguyen T, Mentone S, Thomson B, Biemesderfer D: 174: 5750–5757, 2005 Linking receptor-mediated endocytosis and cell signaling: Evidence 20. Wang Y, Hu Q, Madri JA, Rollins SA, Chodera A, Matis LA: Amelio- for regulated intramembrane proteolysis of megalin in proximal tu- ration of lupus-like autoimmune disease in NZB/WF1 mice after treat- bule. J Biol Chem 279: 34302–34310, 2004 ment with a blocking monoclonal antibody specific for complement 40. Caruso-Neves C, Pinheiro AA, Cai H, Souza-Menezes J, Guggino WB: component C5. Proc Natl Acad Sci U S A 93: 8563–8568, 1996 PKB and megalin determine the survival or death of renal proximal 21. Abbate M, Zoja C, Rottoli D, Corna D, Tomasoni S, Remuzzi G: tubule cells. Proc Natl Acad Sci U S A 103: 18810–18815, 2006 Proximal tubular cells promote fibrogenesis by TGF-␤1-mediated in- 41. Abbate M, Morigi M, Buelli S, Moioli D, Zanchi C, Noris M, Zoja C, duction of peritubular myofibroblasts. Kidney Int 61: 2066–2077, 2002 Remuzzi G: Loss of factor H-dependent complement regulation me- 22. Rangan GK, Pippin JW, Coombes JD, Couser WG: C5b-9 does not diates proximal tubular cells dysfunction on protein overload [Ab- mediate chronic tubulointerstitial disease in the absence of protein- stract]. J Am Soc Nephrol 17: 596A, 2006 uria. Kidney Int 67: 492–503, 2005 42. Abbate M, Zoja C, Morigi M, Rottoli D, Angioletti S, Tomasoni S, 23. Pratt JR, Basheer SA, Sacks SH: Local synthesis of complement com- Zanchi C, Longaretti L, Donadelli R, Remuzzi G: Transforming growth ponent C3 regulates acute renal transplant rejection. Nat Med 8: factor-␤1 is up-regulated by podocytes in response to excess intra- 582–587, 2002 glomerular passage of proteins. Am J Pathol 161: 2179–2193, 2002 24. Eddy AA: Interstitial nephritis induced by protein-overload protein- 43. Turnberg D, Lewis M, Moss J, Xu Y, Botto M, Cook HT: Complement uria. Am J Pathol 135: 719–733, 1989 activation contributes to both glomerular and tubulointerstitial damage in 25. Zhou W, Marsh JE, Sacks SH: Intrarenal synthesis of complement. Adriamycin nephropathy in mice. J Immunol 177: 4094–4102, 2006 Kidney Int 59: 1227–1235, 2001 44. Alexander JJ, Wang Y, Chang A, Jacob A, Minto AW, Karmegam M,

1166 Journal of the American Society of Nephrology J Am Soc Nephrol 19: 1158–1167, 2008 www.jasn.org BASIC RESEARCH

Haas M, Quigg RJ: Mouse podocyte complement factor H: The func- 51. Wada T, Tomosugi N, Naito T, Yokoyama H, Kobayashi K, Harada A, tional analog to human complement receptor 1. J Am Soc Nephrol 18: Mukaida N, Matsushima K: Prevention of proteinuria by the administra- 1157–1166, 2007 tion of anti-interleukin 8 antibody in experimental acute immune com- 45. Hillmen P, Hall C, Marsh JC, Elebute M, Bombara MP, Petro BE, Cullen plex-induced glomerulonephritis. J Exp Med 180: 1135–1140, 1994 MJ, Richards SJ, Rollins SA, Mojcik CF, Rother RP: Effect of eculizumab 52. Ouchterlony O: Diffusion-in-gel methods for immunological analysis. on hemolysis and transfusion requirements in patients with paroxys- II. Prog Allergy 6: 30–154, 1962 mal nocturnal hemoglobinuria. N Engl J Med 350: 552–559, 2004 53. Zhang Z, Schlachta C, Duff J, Stiller C, Grant D, Zhong R: Improved 46. Granger CB, Mahaffey KW, Weaver WD, Theroux P, Hochman JS, techniques for kidney transplantation in mice. Microsurgery 16: 103– Filloon TG, Rollins S, Todaro TG, Nicolau JC, Ruzyllo W, Armstrong 109, 1995 PW: Pexelizumab, an anti-C5 complement antibody, as adjunctive 54. Han WR, Murray-Segal LJ, Mottram PL: Modified technique for kidney therapy to primary percutaneous coronary intervention in acute myo- transplantation in mice. Microsurgery 19: 272–274, 1999 cardial infarction: The COMplement inhibition in Myocardial infarction 55. Ishola DA Jr, van der Giezen DM, Hahnel B, Goldschmeding R, Kriz W, treated with Angioplasty (COMMA) trial. Circulation 108: 1184–1190, Koomans HA, Joles JA: In mice, proteinuria and renal inflammatory 2003 responses to albumin overload are strain-dependent. Nephrol Dial 47. Li JS, Sanders SP, Perry AE, Stinnett SS, Jaggers J, Bokesch P, Reyn- Transplant 21: 591–597, 2006 olds L, Nassar R, Anderson PA: Pharmacokinetics and safety of TP10, 56. van Timmeren MM, Bakker SJ, Stegeman CA, Gans RO, van Goor H: soluble complement receptor 1, in infants undergoing cardiopulmo- Addition of oleic acid to delipidated bovine serum albumin aggra- nary bypass. Am Heart J 147: 173–180, 2004 vates renal damage in experimental protein-overload nephrosis. 48. Benigni A, Corna D, Zoja C, Longaretti L, Gagliardini E, Perico N, Nephrol Dial Transplant 20: 2349–2357, 2005 Coffman TM, Remuzzi G: Targeted deletion of angiotensin II type 1A 57. Benigni A, Zoja C, Tomasoni S, Campana M, Corna D, Zanchi C, receptor does not protect mice from progressive nephropathy of Gagliardini E, Garofano E, Rottoli D, Ito T, Remuzzi G: Transcriptional overload proteinuria. J Am Soc Nephrol 15: 2666–2674, 2004 regulation of nephrin gene by peroxisome proliferator-activated re- 49. Eddy AA, Kim H, Lopez-Guisa J, Oda T, Soloway PD, McCulloch L, Liu ceptor-gamma agonist: Molecular mechanism of the antiproteinuric E, Wing D: Interstitial fibrosis in mice with overload proteinuria: Defi- effect of pioglitazone. J Am Soc Nephrol 17: 1624–1632, 2006 ciency of TIMP-1 is not protective. Kidney Int 58: 618–628, 2000 58. Benigni A, Tomasoni S, Gagliardini E, Zoja C, Grunkemeyer JA, Kalluri 50. Noble B, Olson KA, Milgrom M, Albini B: Tissue deposition of immune R, Remuzzi G: Blocking angiotensin II synthesis/activity preserves glo- complexes in mice receiving daily injections of bovine serum albumin. merular nephrin in rats with severe nephrosis. J Am Soc Nephrol 12: Clin Exp Immunol 42: 255–262, 1980 941–948, 2001

J Am Soc Nephrol 19: 1158–1167, 2008 Filtered Protein Toxicity via C3 1167