Genetics of Immune-Mediated Glomerular Diseases (PDF)

Genetics of Immune-Mediated Glomerular Diseases: Focus on Complement Marina Noris, PhD,* and Giuseppe Remuzzi, MD*,†,‡ Summary: The spectrum of immune-mediated glomerular diseases is wide, ranging from rare disea...

Genetics of Immune-Mediated Glomerular Diseases: Focus on Complement Marina Noris, PhD,* and Giuseppe Remuzzi, MD*,†,‡ Summary: The spectrum of immune-mediated glomerular diseases is wide, ranging from rare diseases with well-recognized genetic origins to more common and multifactorial diseases. Immune-mediated glomerular injury is complex and involves both the innate and the adaptive immune systems. In the past 20 years a huge effort has been undertaken to unravel the genetic basis of immune-mediated glomerular diseases. The discovery of abnormalities in genes encoding proteins of the alternative pathway of complement in more than 50% of patients with atypical hemolytic uremic syndrome (aHUS), and in approximately 20% of patients with membranoproliferative glomerulonephritis (MPGN), has highlighted the role of this complement pathway in the pathogenesis of immune-mediated glomerular diseases. aHUS-associated complement gene abnormalities mainly result in complement dysregulation restricted to the cell surface, whereas complement activation in the fluid phase prevails in most, but not all, genetic cases of MPGN. Results achieved in aHUS and MPGN have boosted interest in the impact of complement gene abnormalities and variations in the predisposition to more common, multifactorial kidney diseases, including IgA nephropathy and lupus nephritis. Emerging findings in these complex diseases have broadened our understanding of the fragile balance between the protective and harmful functions of the complement system. Semin Nephrol 37:447-463 C 2017 Elsevier Inc. All rights reserved. Keywords: Genetics, glomerular diseases, atypical hemolytic uremic syndrome, complement, membrano- proliferative glomerulonephritis, lupus nephritis, IgA nephropathy I mmune-mediated damage to glomerular structures (GBM) antibody disease, membranous nephropathy, largely is responsible for the pathogenesis of the and many others.1 Immune-mediated glomerular injury majority of glomerular diseases. These include is complex and implies the activation of both the innate atypical hemolytic uremic syndrome (aHUS), mem- and the adaptive immune system. Many immune- branoproliferative glomerulonephritis (MPGN), IgAN, mediated glomerular diseases are rare and disease lupus nephritis (LN), acute poststreptococcal glomer- clusters have been reported often in families, suggest- ulonephritis, anti–glomerular basement membrane ing the existence of genetic determinants that may contribute to disease risk. In the past 20 years a huge effort has been made to unravel the genetic basis of * IRCCS, Istituto di Ricerche Farmacologiche “Mario Negri,” immune-mediated glomerular diseases based on direct Clinical Research Center for Rare Diseases “Aldo e Cele Daccò,” Ranica, Bergamo, Italy. sequencing of candidate genes, whole-genome linkage † analysis in families, and genome-wide association Unit of Nephrology and Dialysis, Azienda Ospedaliera Papa Giovanni XXIII, Bergamo, Italy. studies (GWAS). Hundreds of risk variants in genes ‡ Department of Biomedical and Clinical Sciences, University of of the HLA system and of the innate and adaptive Milan, Milan, Italy. immune response have been reported. Financial support: This work was supported by Fondazione ART The discovery that mutations in genes encoding per la Ricerca sui Trapianti ART ONLUS (Milan, Italy) and by the FP7 EU project (EURenOmics project, HEALTH-F5-2012- proteins of the alternative pathway of complement are 305608). associated with two rare glomerular diseases, aHUS2 Conflict of interest statement: Marina Noris has received honoraria and MPGN,3 has highlighted the importance of genetic from Alexion Pharmaceuticals for giving lectures and participat- variations in the complement system in determining ing in advisory boards and research grants from Omeros and predisposition to glomerular injury, and encouraged Chemocentryx. Giuseppe Remuzzi has consultancy agreements research on complement genetics in more complex with AbbVie*, Alexion Pharmaceuticals*, Bayer Healthcare*, Reata Pharmaceuticals*, Novartis Pharma*, AstraZeneca*, multifactorial immune-mediated glomerular diseases Otsuka Pharmaceutical Europe*, Concert Pharmaceuticals*. such as IgAN and LN. In this article we provide an None of these activities has had any influence in this article. (*No overview of the known and emerging genetic evidence personal remuneration is accepted, compensations are paid to his regarding the role of genetic variations in complement institution for research and educational activities.). genes in the pathogenesis of these diseases. Address reprint requests to Giuseppe Remuzzi, MD, IRCCS, Mario Negri Institute for Pharmacological Research, Clinical Research Center for Rare Diseases “Aldo e Cele Daccò,” Via Camozzi 3, 24020 Ranica Bergamo, Italy. E-mail: giuseppe. THE COMPLEMENT SYSTEM [email protected] 0270-9295/ - see front matter The complement system is part of innate immunity and & 2017 Elsevier Inc. All rights reserved. functions as a first-line defense against pathogens, but http://dx.doi.org/10.1016/j.semnephrol.2017.05.018 also plays a central role in the clearance of immune Seminars in Nephrology, Vol 37, No 5, September 2017, pp 447–463 447 448 M.R. Noris and G. Remuzzi Classical pathway Lectin pathway Alternative pathway IgM, IgG Mannose bacteria,viruses, tick-over Immune complexes residues C1q,C1r,C1s MBL, Ficolins FB FD C3(H2O)Bb MASP1-3 C3a FI FH, MCP, MCP, CR1 C4,C2 C3b C1inh C4BP FI C1inh CR1, THBD C4BP, DAF C4bC2a FB FD C3bBb FH, CR1, C3 convertase DAF C3 convertase P C3 C3 C3a C3a C3b C3b C4bC2aC3b (C3b)2Bb FH C5 convertase C5 convertase C5 C5a CD59 C5b-9 Clusterin Vitronectin Figure 1. The three complement pathways. The classic pathway is activated by the binding of the Fc region of IgG or IgM antibodies to the complement complex C1, comprising one C1q molecule, two C1r molecules, and two C1s molecules. The lectin pathway is triggered by the binding of MBL or ficolins to sugar molecules on pathogens or altered self, which leads to the activation of MBL-associated serine proteases (MASP-1, MASP-2, and MASP-3). The classic and lectin pathways converge into the cleavage of complement components C2 and C4, leading to the formation of the C3 convertase (C4bC2a) of the classic/lectin pathways. The alternative pathway is activated continuously in plasma by low-grade hydrolysis (tick-over) of C3 that forms C3(H2O). The latter binds to FB, which in turn is cleaved by factor D (FD) to form the alternative pathway fluid-phase C3 convertase. The C3 convertases cleave C3 into C3a, an anaphylotoxin, and C3b that deposits on cell surfaces. C3b contributes to the formation of the alternative pathway surface C3 convertase that cleaves additional C3 molecules, resulting in an amplification loop. In addition, C3b contributes to the formation of the C5 convertases that cleave the complement component C5, producing the anaphylatoxin C5a and C5b. C5b initiates the late events of complement activation, leading to the formation of the membrane-attack complex (C5b-9 complex). Self-surfaces are protected against complement damage by protein regulators (red characters). C1inh, C1 inhibitor, inactivates C1r and C1s, MASP-1, and MASP- 2; C4BP, C4b-binding protein, binds to C4b and has decay accelerating activity for the classic pathway C3 convertase and cofactor activity for factor I–mediated C4b cleavage; CD59, protectin, vitronectin, and clusterin, prevents C5b-9 formation; CR1, has decay accelerating activity as well as cofactor activity for factor I–mediated C3b and C4b cleavage; DAF, has decay accelerating activity on C3/C5 convertases of the classic and alternative pathways; FH, binds C3b, exerts cofactor activity for factor I–mediated C3b cleavage, and dissociates (decay accelerating activity) the alternative pathway C3 and C5 convertases; MCP, has cofactor activity for factor I– mediated C3b and C4b cleavage; FI, it degrades C3b and C4b aided by cofactors. complexes (ICs) and damaged cells, and in the regu- favor the removal of target cells or IC by leukocytes; lation of adaptive immune response.4 Complement and (3) the terminal C5b-9 complex that lyses patho- components and regulators are organized into three gens or damaged self-cells (Fig. 1). activation pathways: the alternative, classic, and lectin The classic pathway is triggered by interaction pathways, the activation of which results in the between C1q and antigen–antibody IC in the circula- formation of C3 convertases, which are protease tion or on target cells, whereas the lectin pathway uses complexes that cleave C3 into C3a and C3b. The mannose binding lectins (MBLs) and ficolins to iden- binding of additional C3b to the C3 convertases forms tify carbohydrate ligands on the surface of microbes or the C5 convertases that cleave C5, producing C5a and damaged host cells.4 C5b. At this stage the three pathways converge into a The alternative pathway always is activated at a low common terminal pathway that leads to the formation degree in fluid phase by hydrolysis of C3, which of the terminal complement complex C5b-9 (Fig. 1). through the formation of the alternative pathway Activated complement generates three major groups initiation C3 convertase leads to the deposition of of effector molecules: (1) anaphylatoxins (C3a and low amount of C3b onto cell surfaces. C3b binds C5a), which attract and activate leukocytes through factor B (FB), and forms the surface alternative path- interaction with C3a and C5a receptors; (2) opsonins way C3 convertase (Fig. 1) that cleaves additional C3 (C3b, inactivated C3b [iC3b], and C3d), which cova- molecules. C3b generated by any of the three pathways lently bind to target surfaces to facilitate transport and can form the alternative pathway C3 convertase, which Immune-mediated glomerular diseases 449 serves as an amplification loop for the entire comple- patients show normal circulating C3 levels.5 In the past ment system. For this reason, the complement system 20 years genetic studies have provided convincing is finely regulated by a set of molecules that are either evidence that aHUS is linked to genetically determined membrane-anchored or plasmatic, so as to protect self- dysregulation of alternative pathway of complement. surfaces from complement damage (Fig. 1). Regulatory Functional studies have shown that aHUS-associated proteins prevent the formation of the C3 convertase, mutations mainly result in complement dysregulation foster iC3b inactivation by factor I (FI) (cofactor that is restricted to cell surfaces,5 and proceeds until the activity), dissociate C3- and C5-convertases (decay terminal pathway with the formation of C5b-9. In acceleration activity), or prevent C5b-9 complex assem- contrast, the regulation of the alternative pathway often bly (Fig. 1). The perturbation of the balance between is normal in the fluid phase.5 complement activators and regulators provides the C5b-9 deposition alters the endothelial cell pheno- molecular basis of several glomerular diseases. type profoundly, causing the expression of tissue factor and adhesion molecules and cell retraction, thus exposing the underlying prothrombotic matrix.7 ATYPICAL HEMOLYTIC UREMIC SYNDROME Together, these events activate platelets and the coag- ulation system. The pathogenetic role of the terminal Clinical Features, Histology Features, and Pathogenesis complement pathway in aHUS was confirmed even- HUS is defined clinically by thrombocytopenia, non- tually by the effectiveness of the anti-C5 antibody immune microangiopathic hemolytic anemia (Coombs eculizumab in inducing disease remission and in negative), and acute kidney injury. Markers of intra- protecting patients from microvascular thrombosis.8 vascular hemolysis include increased plasma lactate dehydrogenase levels and undetectable or reduced Genetics plasma haptoglobin level.2 The underlying histology lesion, defined as thrombotic microangiopathy, affects Complement factor H mutations, complement factor arterioles and capillaries in the vasculature of the H–complement factor H-related hybrid genes, and kidney and other organs including the brain, heart, anti–factor H autoantibodies lung, gut, pancreas, and liver, and is characterized by a The most important regulator of the alternative path- thickening of the vascular walls with prominent endo- way, complement factor H (CFH), is composed of 20 thelial swelling and detachment and accumulation of short consensus repeats (SCRs). FH regulates the proteins and cell debris in the subendothelium. In the alternative pathway both in the fluid phase and on cell kidney the lesions are observed mainly in glomerular surfaces. The first four SCRs mediate the regulatory capillaries and preglomerular arteries, but arterioles functions (cofactor activity and decay accelerating and interlobular arteries also may be involved.2 activity), whereas the two carboxy-terminal SCRs 19- Thrombocytopenia reflects the formation of platelet- 20 form a recognition domain for polyanions and C3 rich and fibrin-rich thrombi in the microcirculation activation compounds on host cell surfaces. Mutations that, together with endothelial swelling, obstruct vessel in CFH are the most frequent genetic abnormalities in lumina, leading to organ ischemia and dysfunction. aHUS patients, accounting for 20% to 30% of cases9,10 Anemia is thought to be caused by red blood cells that (Table 1). A few patients have homozygous mutations break apart as they squeeze past the obstructed small and low FH levels. However, most mutations are blood vessels, but complement-mediated hemolysis heterozygous, are associated with normal FH levels, and accelerated removal of damaged erythrocytes in affect the carboxy-terminal SCRs, and result in a the spleen also may play a role. mutant protein that is unable to bind to and regulate Most cases (490% in children) are associated with complement on endothelial cells and platelets.9–11 infections by Shiga-like toxin-producing bacteria. The The CFH gene is located close to the CFH-related term aHUS has been used to describe those rare cases (CFHR)1-5 genes encoding five FH-related proteins (o10%) in which infections by Shiga-like toxin-produc- (Fig. 2A). CFH and CFHRs share a high degree of ing bacteria, Streptococcus pneumoniae, or other secon- sequence identity that predisposes them to gene con- dary causes can be excluded. The incidence of aHUS is versions and genomic rearrangements. Hybrid CFH/ approximately 0.5 to 1 people per million per year.2 CFHR1 and CFH/CFHR3 genes coding abnormal FH It was suspected that the activation of the alternative proteins in which the carboxy-terminal SCRs are sub- pathway of complement plays a role in the patho- stituted for those of FHR1 or by the entire FHR3, genesis of aHUS because C3 deposits have been found resulting in hybrid FH molecules with decreased comple- along the glomerular capillaries and on the endothe- ment regulatory activity on endothelial surfaces (Fig. 2B), lium of arterioles and small arteries of patients.2,5 have been reported in 3% to 5% of aHUS patients.12,13 Reduced serum C3 levels with normal C4 levels also An aHUS-associated reverse hybrid CFHR1/CFH gene were reported,6 but approximately 40% to 50% of also has been reported, and it encodes a FHR1 protein in 450 M.R. Noris and G. Remuzzi Table 1. Complement Gene Abnormalities Associated With aHUS and IC-MPGN/C3G aHUS Gene Abnormality Main Effect Frequency CFH Heterozygous and (rarely) Impaired cell-surface complement regulation 20%-30% homozygous mutations mainly in the last two exons CFH/CFHRs Nonallelic homologous recombinations Impaired cell-surface complement regulation 3%-5% CFHR1/CFHR3 Homozygous deletion and formation of Impaired cell-surface complement regulation 5%-10% CFHR1/CFHR4 anti-FH antibodies MCP Heterozygous and (rarely) Reduced cell-surface expression 8%-10% homozygous mutations CFI Heterozygous mutations Reduced secretion, low co-factor activity 4%-8% C3 Heterozygous mutations Decreased C3b inactivation 4%-8% CFB Heterozygous mutations C3 convertase stabilization 1%-3% THBD Heterozygous mutations Reduced cell-surface C3b inactivation 3%-4% DGKE Homozygous or compound Infantile heterozygous mutations cases* Combined Heterozygous mutations in two or Impaired cell-surface complement regulation 3%-4% mutations three complement genes IC-MPGN/C3G CFH Heterozygous and homozygous Impaired fluid phase complement regulation 4%-16% mutations mainly in first exons CFHRs/CFHRs Nonallelic homologous recombinations Increased competition with FH Rare MCP Heterozygous mutations Not yet studied 0%-1% CFI Heterozygous mutations Not yet studied 2%-5% C3 Heterozygous mutations C3 convertase resistant to FH-mediated 9%-13% decay, reduced CR1-dependent C3b degradation CFB Heterozygous and (rarely) Not yet studied homozygous mutations *Onset of the disease before 1 year of age. which SCR5 of FHR1 has been replaced by SCR20 of Membrane cofactor protein mutations FH, and impairs complement regulation by competing Membrane cofactor protein (MCP) encodes the surface with binding of FH to endothelial cells14 (Fig. 2B). complement regulator MCP, which has cofactor activ- Anti-FH autoantibodies are found in 5% to 10% of ity for the cleavage of C3b and C4b by FI (Fig. 1). aHUS patients15 (Table 1) and in approximately 25% MCP mutations, most of them heterozygous, have been to 50% of pediatric cases. The antibodies bind prefer- identified in 8% to 10% of patients. Most patients entially to the carboxy-terminal SCRs of FH, and in (
75%) have decreased MCP expression on blood 77.4% of samples they react with SCRs 19-20.15 The leukocytes. Less frequently, MCP expression is nor- functional consequences of the antibodies mimic those mal, but the protein is dysfunctional (Table 1).17,18 of CFH genetic defects; namely, reduced FH binding to C3 fragments on the cell surface, resulting in impaired Complement factor I mutations FH-mediated cell surface protection (Table 1).15 The development of anti-FH antibodies has a genetic Mutations in the gene encoding FI, the plasma protease predisposition, being associated strongly with the that cleaves C3b and C4b (Fig. 1), have been found in homozygous deletion of CFHR1 and CFHR316 4% to 8% of patients (Table 1). Eighty percent cluster (Table 1). In a few patients the anti-FH antibodies in the serine–protease domain. Approximately 50% of are associated with CFHR1 mutations or with a mutants are not secreted, although some mutants are deletion including CFHR1 and CFHR4, suggesting secreted but have impaired proteolytic activity.19,20 that the FHR1 deficiency is the predominant predis- posing factor for the development of these antibodies. C3 and complement FB mutations However, FHR1 deficiency per se is not enough to Gain-of-function mutations, mostly heterozygous, in cause anti-FH antibodies and aHUS, because the the genes encoding the two components of the alter- homozygous CFHR1 and CFHR3 deletion is present native pathway C3 convertase, C3 and FB (Fig. 1), in 3% to 6% of Europeans and in approximately 20% account for 4% to 8% and 1% to 3% of aHUS cases, of the African population. respectively (Table 1).9,21,22 Most C3 mutations result Immune-mediated glomerular diseases 451 A FH 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 FHR1 1 2 3 4 5 FHR2 1 2 3 4 FHR3 1 2 3 4 5 FHR4 1 2 3 4 5 6 7 8 9 FHR5 1 2 3 4 5 6 7 8 9 B aHUS FH/FHR1 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 4 5 FH/FHR1 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 5 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 1 2 3 4 5 FH/FHR3 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 1 2 3 4 5 FH/FHR3 FHR1/FH 1 2 3 19 20 FHR1/FH 1 2 3 4 20 C3G 1 2 1 2 1 2 3 4 5 FHR112341234 FHR2/FHR5 1 2 1 2 3 4 5 6 7 8 9 FHR3/FHR1 1 2 1 2 3 4 5 1 2 FHR5/FHR2 1 2 3 4 1 2 1 2 3 4 6 7 8 9 FHR51212 5 Figure 2. Structures of (A) normal FH and FHR and of (B) hybrid proteins found in patients with aHUS or C3G. SCRs are represented by ovals and are numbered from the N-terminal end. Homologous SCRs are aligned. The SCRs of the FH regulatory domain are shown in blue, the SCRs of the surface recognition domain are in red. in defective binding of C3b to MCP and/or FH, which of patients with aHUS (Table 1).9,25 aHUS-associated translates to decreased inactivation of C3b and mutations impair C3b inactivation on the cell surface.25 increased deposition on endothelial cells.23 Some complement FB (CFB) mutations result in super-B, Combined genetic and environmental factors contribute to which forms a hyper-reactive C3 convertase that aHUS is resistant to dissociation by complement regulators.22,24 However, half of the CFB mutations Mutations in complement genes can be found in healthy did not result in any FB functional abnormality family members. Indeed, incomplete penetrance of in vitro,24 suggesting that these variants are unrelated aHUS has been reported in approximately 50% of to disease pathogenesis. individuals carrying mutations in CFH, complement factor I (CFI), MCP, CFB, and C3.9,26 In addition, the age of onset and severity of disease vary greatly among Thrombomodulin mutations patients with the same mutations.26,27 This indicates Heterozygous mutations in the thrombomodulin gene that penetrance and phenotype are determined by other (THBD) encoding thrombomodulin, an endothelial genetic and environmental modifiers. anticoagulant protein that also regulates the alternative Approximately 3% to 4% of patients carry muta- pathway of complement, have been found in 3% to 4% tions in more than one complement gene, or mutations 452 M.R. Noris and G. Remuzzi Table 2. Risk and Protective SNPs and Haplotypes for aHUS and IC-MPGN/C3G Gene Variant (SNP, rs) Effect Haplotype Impact Disease CFH -332T (c.-332C4T, rs3753394 ) Unknown Risk aHUS 184G (c.184G4A, p.V62I, rs 800292) Reduced cofactor activity Risk C3G (DDD, C3GN) 1204C (c.1204T4C, p.Y402H, rs1061170) Reduced affinity to certain Risk DDD surface molecules 2016G (c.2016A4G, p.Q672Q, rs3753396) Unknown Risk aHUS 2808T (c.2808G4T, p.E936D, rs1065489) Unknown Risk aHUS H1 (CGCAG) Risk DDD H2 (CATAG) Protective aHUS, DDD H3 (TGTGT) Risk aHUS CFHR1 469T (c.469C4T, p.H157Y, rs425757) Competition with FH at 475G (c.475C4G, p.L159V, rs410232) cell surface level 523C (c.523G4C, p.E175Q, rs388862) CFHR1*B (TGC) Risk aHUS MCP -652A (c.-652A4G, rs2796267) Unknown Risk IC-MPGN, C3GN -366A (c.-366A4G, rs2796268) Unknown Risk IC-MPGN IVS9-78G4A (rs1962149) Unknown Unknown IVS12þ638G4A (rs859705) Unknown Unknown c.*783T (c.*783T4C, rs7144) Unknown Risk IC-MPGN MCP1 (AAGGT) Risk IC-MPGN, C3GN MCP2 (GGAAC) Risk aHUS MCP3 (GAGGT) Protective C3G (C3GN, DDD) C3 304G (c.304C4G, p.R102G, rs2230199) Unknown Risk DDD 941T (c.941C4T, p.314L, 1047286) Unknown Risk DDD THBD 1418C (c.1418C4T, p.A473V, rs 1042579) Unknown Risk C3G in one complement gene plus anti-FH autoantibodies that aHUS is the result of an otherwise innocuous (Table 1).28,29 stimulus that triggers the alternative pathway and sets Single-nucleotide polymorphisms (SNPs), defined off a self-amplifying cycle that cannot be controlled as genetic changes with more than a 1% minor allele appropriately in genetically susceptible individuals. frequency in the normal population, have been asso- ciated with aHUS.30 A CFH haplotype (CFH-H3) Diacylglycerol kinase ε mutations (Table 2) including five SNPs in the promoter and in Recently, homozygous or compound heterozygous coding regions, is associated with a two- to four-fold mutations in diacylglycerol kinase ε (DGKE), encod- increased risk of aHUS.31 One of the variants in this ing the intracellular protein DGKE, have been reported haplotype, FH p.Val 62, has decreased cofactor activity in a recessive pediatric form of aHUS (Table 1).35,36 compared with the protective Ile variant.32 Another Mutation carriers presented with aHUS before 1 year aHUS risk haplotype has been reported in MCP of age with hypertension, hematuria, and severe (MCPGGAAC, MCP2) (Table 2), including two SNPs proteinuria. DGKE is a component of a lipid kinase in the promoter.33 A CFHR1 haplotype variant family that terminates diacylglycerol signaling. It is not (CFHR1*B) is associated with aHUS in homozygous related directly to complement. DGKE silencing in carriers (Table 2).34 It has been suggested that because endothelial cells induced a proinflammatory and pro- SCR3 of FHR1*B is identical to SCR18 of FH, it may thrombotic phenotype, and increased endothelial apop- compete with FH at the endothelial cell level. tosis,37 thus recapitulating the pathogenetic effects of These haplotypes and SNPs act together with complement gene mutations. Interestingly, combined mutations to increase susceptibility to the disease. mutations in DGKE and complement genes have been Penetrance increased as the number of genetic alter- reported recently in a few pediatric patients.38 ations in a patient increased.28 Even in subjects with multiple genetic/acquired risk factors, aHUS may not occur until middle age, sug- MEMBRANOPROLIFERATIVE gesting that a triggering stimulus is required for the GLOMERULONEPHRITIS disease to manifest. Precipitating events, most commonly upper respira- Clinical Features, Histology Features, and Pathogenesis tory tract infections or gastroenteritis, but also preg- MPGN is a kidney disorder that affects mainly children nancy, tissue and organ ischemia, and drugs, have been and young adults. The clinical manifestations are reported in more than half of patients.9 Thus, it is likely variable, from isolated hematuria and proteinuria to Immune-mediated glomerular diseases 453 nephrotic or nephritic syndrome or rapidly progressive However, a number of findings challenge the glomerulonephritis.39 Approximately 50% of patients earlier-described, IF-based classification. Glomerular develop end-stage renal failure within 10 to 20 years of Ig staining is positive in a substantial number of diagnosis. Hypocomplementemia (low C3 level) is a patients with DDD and C3GN, so that 10% to 20% very common finding. MPGN may be secondary to of DDD cases and a higher percentage of C3GN infections autoimmune diseases and malignancies, or patients would not be classified as having C3G idiopathic when a clear underlying etiology cannot be according to the IF pattern.40 Conversely, isolated identified.39 Idiopathic or primary MPGN is rare, with staining for C3 on IF may be observed in cases of a prevalence in the European Union of approximately postinfectious glomerulonephritis. Analysis of 6 cases per 100,000 people. repeated biopsies has shown a change in IF pattern The pattern of glomerular injury by light micro- in 10 of 23 biopsy specimens, with some patients scopy typically includes mesangial hypercellularity and showing a pattern of IC-MPGN in the first biopsy matrix expansion, thickening of glomerular capillaries and C3G on a repeat biopsy, and vice versa for other with the formation of double contours, interposition of patients.43 Finally, alternative pathway abnormalities leukocytes and mesangial cells, and synthesis of new with low serum C3 levels with normal C4 levels, GBM.39 These changes are the result of deposition of C3 nephritic factor, and/or mutations in complement IC and complement factors in the mesangium and genes were found as frequently in patients along the capillary walls. with primary IC-MPGN as in patients with C3G, Traditionally, through electron microscopy, MPGN indicating that IC-MPGN and C3G may have more was classified into type I, with subendothelial deposits; commonalities than previously recognized, with dys- type II or dense deposit disease (DDD), with intra- regulation of the alternative pathway playing a main membranous highly electron-dense deposits; and type pathogenetic role.41,44 III, with subendothelial and subepithelial deposits.39 A Emerging studies on the genetic determinants of IC- more recent classification is based on the compositions MPGN and C3G will contribute to a new, more of the deposits, centered on whether or not Ig deposits sophisticated classification of these diseases that hope- accompany C3 on immunofluorescence (IF).39 MPGN fully will include novel discoveries on the genetic basis associated with substantial Ig deposits has been termed of diverse disease manifestations. Repeated biopsies to immune-complex–mediated MPGN (IC-MPGN) and is look for and quantify Ig and complement deposits associated commonly with malignancies, chronic carefully, and the evaluation of complement activation infections, or autoimmune diseases, although in some markers in serum (C3) and plasma (sC5b-9) are IC-MPGN cases an underlying cause cannot be iden- suggested to disclose immune-mediated forms and tified.3 It has been proposed that the activation of the distinguish them from cases dependent on alternative complement classic pathway plays a role in IC-MPGN pathway of complement activation in the fluid phase or because of antigen-antibody IC caused by chronic viral on the cell surface, which may benefit from comple- and bacterial infections, increased levels of circulating ment inhibition treatment. IC, or paraproteinemias.39 MPGN cases that present with predominant C3 Genetics glomerular staining are considered complement- mediated and called C3 glomerulopathy (C3G).40 On Complement gene mutations the basis of electron microscopy findings, C3G is The reports of familial cases of IC-MPGN and C3G45 divided further into DDD, with highly electron-dense have indicated the diseases had a genetic basis. deposits; and C3 glomerulonephritis (C3GN), with Complete deficiency of FH, associated with homozy- mesangial, subendothelial, and sometimes subepithelial gous or compound heterozygous CFH mutations, have deposits. Thus, C3GN includes those examples of been described in patients with either IC-MPGN or MPGN type I and III in which IF shows predominant C3G,46–48 with most developing renal disease during C3 deposits. The dysregulation of the alternative path- infancy. Subsequently, heterozygous CFH mutations way of complement appears to be a primary cause of have been reported both in adult and childhood C3G, and may be associated with C3 nephritic factors, cases,41,44,49 indicating that FH haplodeficiency also autoantibodies that stabilize the C3 convertase of the may predispose to these diseases (Table 1). Recent alternative pathway, or with defects in genes encoding screening in a large cohort of patients with primary complement proteins,40–42 resulting in fluid-phase forms of IC-MPGN or with C3G identified mutations complement activation. The transfer of circulating in complement regulatory genes (CFH, CFI, and MCP) complement products, such as C3b and C5b-9, through and in the genes encoding C3 and FB in 24 of 140 the fenestrated glomerular endothelial cells to the patients.44 Notably, the prevalence of mutations was subendothelium and mesangium, triggers glomerular comparable between IC-MPGN and C3G patients inflammation and a proliferative response. (17% and 18%, respectively).44 Similar results were 454 M.R. Noris and G. Remuzzi reported previously in a French cohort,41 showing a FH-mediated decay. Decay of the mutant C3 conver- similar prevalence of mutations in CFH, CFI, or MCP tase by the surface regulator decay accelerating factor, —C3 and CFB were not analyzed in this study— and C3b inactivation in the presence of FI and MCP among patients with C3G (19%) and patients with were normal, indicating that the C3 mutation causes MPGN 1 (17%), the latter group presenting a pattern of fluid phase–restricted alternative pathway dysregula- IC-MPGN at IF (Table 1). tion (Table 1). This finding would support the hypoth- It is plausible that in some cases of IC-MPGN an esis that continuous C3b generation in plasma via the inciting autoimmune or infection event initially acti- alternative pathway plays a role in DDD. On the other vated the classic pathway, and then chronic comple- hand, a heterozygous p.I734T C3 mutation found in ment activation was sustained by genetic or acquired two brothers with C3GN and normal C3 levels caused abnormalities leading to alternative pathway dysregu- the formation of a C3 convertase that normally was lation and chronic kidney disease.50 regulated by FH.54 However, this mutant showed IC-MPGN/C3G-associated mutations affect the decreased binding to complement receptor 1 (CR1), a same genes involved in aHUS, but there are important complement regulator that is expressed highly on differences. Mutations affecting C3 and CFB are the podocytes, resulting in less CR1-dependent cleavage most prevalent in patients with IC-MPGN and C3G, of C3b by FI. Increased deposition of C3 was observed followed by CFH mutations,44,51 unlike with aHUS- on cultured podocytes and glomerular endothelial cells associated mutations, which most commonly affect after incubation with patients’ serum, compared with CFH (Table 1).9 IC-MPGN/C3G-associated CFH normal serum (Table 1).54 This evidence highlights the mutations usually affect the amino-terminus of the role of podocytes as targets of complement in this protein, and either cause premature truncation or disease. impaired protein secretion, resulting in low or unde- tectable FH levels,45 or more rarely result in normally secreted FH protein with severely reduced cofactor and Risk factors decay-accelerating activities in fluid phase.47 The Polymorphic variants and haplotypes in complement absence of FH function in plasma causes massive genes (Table 2) have been shown to influence suscept- fluid-phase alternative pathway activation and con- ibility to IC-MPGN or C3G. sumption of complement components, reflected by The C variant of the CFH c.1204T4C SNP that low serum C3 and C5 levels. At variance with this, leads to a substitution of tyrosine 402 with a histidine, aHUS-associated CFH mutations are mostly hetero- is over-represented in DDD patients compared with zygous and affect the carboxy-terminus of FH, impair- controls.41,55 In another study, C3G patients (including ing its capacity to protect endothelial cells from DDD and C3GN) had a higher frequency of the complement attack, leading to surface-restricted com- common G allele (encoding a valine) of another CFH plement dysregulation.2,5 However, there are excep- SNP, c.184G4A (p.V62I)44 (Table 2). Both disease tions. The p.R1210C CFH mutation is the most risk SNPs have functional effects; indeed, the H402 frequent genetic abnormality in aHUS,9 but also has variant shows a reduced affinity for certain surface been found in association with C3G.44 Functional molecules,56 whereas the Val62 CFH showed lower studies have shown that this mutation generates cova- binding affinity for C3b and reduced cofactor activity lent complexes between FH and albumin that impair than the 62Ile variant.32 Interestingly, the earlier- accessibility to all FH functional domains, pre- described two risk alleles are part of the H1 CFH disposing individuals to various pathologies.52 The haplotype that was associated with DDD,31,41 and are final disease phenotype is determined by other absent in the DDD protective CFH H2 haplotype genetic risk factors. Conversely, few patients with (Table 2).55 aHUS have mutations affecting the FH amino- The A alleles of the c.-652A4G and c.-366A4G terminus. These data indicate a link in the spectrum SNPs in the MCP promoter have been found to be of kidney diseases determined by alternative pathway over-represented in patients with C3GN or IC-MPGN dysregulation, as supported by data from patients with (Table 2).41,44 Interestingly, both alleles are part of a a mixed histology pattern of MPGN and HUS on the risk haplotype (MCP1) (Table 2) for C3GN and same biopsy or on biopsy specimens taken at different MPGN type 1.41 Another C3G risk allele was found time points.9 in the common c.1418C4T (p.A473V) SNP of Only two IC-MPGN/C3G-associated C3 mutants THBD, with C3G patients showing a higher frequency have been characterized and no functional studies have of the A473 variant compared with controls been conducted for FB mutants. A DDD-associated (Table 2).44 Thrombomodulin is an anticoagulant heterozygous C3 mutation results in a C3 protein endothelial glycoprotein that also regulates the comple- that generates a hyperfunctional C3 convertase that ment alternative pathway.57 Increased thrombomodulin cleaves circulating wild-type C3,53 and is resistant to expression on glomerular endothelial cells was found Immune-mediated glomerular diseases 455 in biopsy specimens of patients with MPGN, suggest- the avidity of FHR1, FHR2, and FHR5 for their ligands ing that thrombomodulin may be important to counter- C3b, iC3b, and C3dg, and enhances the ability of FHR1 balance complement hyperactivation occurring in and FHR5 to compete with FH for the earlier-described MPGN glomeruli. ligands.65 In patients with C3G and CFHR genomic Finally, the G variant of the p.R102G and the L abnormalities, the duplication of the dimerization motif variant of the p.P314L SNPs of C3 have been reported increases oligomerization of FHR1, FHR2, and to confer susceptibility to DDD (Table 2).55 FHR5.60 Because FHR proteins are either devoid of or Combinations of risk and protective variants and have much lower complement regulatory activity than haplotypes in complement genes may determine the FH,64 it was proposed that increased oligomerization levels of alternative pathway activity in the fluid phase results in gain-of-function proteins that overcompete or on cell surfaces and, in the presence of other genetic with FH binding to C3b and host surface and impair and acquired risk factors, may influence disease risk. complement regulation,64 a phenomenon called de- This possibility has been highlighted recently by regulation. Functional studies showing that serum frac- association studies, showing that the risk of developing tions containing FHR proteins from patients with either IC-MPGN or C3G strongly increases in subjects the FHR5 or the FHR1 internal duplications,60,65 or the carrying complement gene mutations combined with FHR3/FHR1 or FHR5/FHR2 hybrid proteins,63 common susceptibility variants (FH p.V62 and THBD enhanced guinea pig erythrocyte cell lysis, are consis- p.A473 for C3G or MCP c.-366A for IC-MPGN),44 tent with surface FH de-regulation. compared with subjects with mutations but no suscept- Notably, C3G-associated genomic rearrange- ibility variant. ments differ from those associated with aHUS because Notably, IC-MPGN/C3G risk variants/haplotypes the latter always involve CFH and cause the exchange are different from those associated with aHUS and of the carboxy–terminal regions between FH and such modifiers may determine the kidney phenotype. FHR1 (Table 1 and Fig. 1B), resulting in impaired In a report on two unrelated patients with complete FH complement regulation on the endothelial cell deficiency, one patient who also was homozygous for surface.64 the MCP1 IC-MPGN/C3G risk haplotype developed A possible explanation for why the C3G-associated endocapillary glomerulonephritis with C3 deposits, and FHR hybrids do not cause aHUS could be that they the other patient with the homozygous aHUS MCP2 compete with FH only on a subset of surfaces that does risk haplotype had aHUS.58 not include endothelial cells. The GBM is a good candidate surface for FH/FHR competition because FH, FHR1, and FHR5 bind components of the extrac- CFHR genomic rearrangements ellular matrix,64 and FHR5 inhibits the surface cofactor Complex genomic abnormalities have been reported in activity of FH by competing with its binding to the C3G, which affect the CFHR genes, resulting in extracellular matrix. This competition may increase in internal duplications or formation of FHR hybrids presence along the GBM of C3 activation fragments, (Fig. 2B and Table 1). Among a large cohort with bacterial products, properdin, or apoptotic cells, which Greek Cypriot ancestry, a form of C3G, also known as are recognized by both FH and FHRs.64,66 In this CFHR5 nephropathy, segregated with a duplication of regard, dimerization has been shown to enhance the exons two to three of CFHR5, which encode the first interaction of human FHR5 with GBM-bound C3 in two N-terminal SCRs.59 The same mutant FHR5 mouse glomeruli.65 protein caused by a distinct DNA breakpoint was Most patients with C3G and abnormal FHRs have described in a family with C3G without Cypriot intense C3 staining but normal or only moderately ancestry. A duplication of the first four SCRs of reduced C3 levels, a finding consistent with FHR1 was reported in affected members of a family glomerular-restricted complement de-regulation. How- with autosomal-dominant C3G (Fig. 2B).60 Genomic ever, the FHR2/FHR5 hybrid protein, reported in two rearrangements among different CFHR also have been siblings with DDD,62 was associated with very low C3 found in C3G pedigrees, resulting in CFHR3/ levels and inhibited the dissociation of the alternative CFHR1,61 CFHR2/CFHR5,62 and CFHR5/CFHR263 pathway C3 convertase by FH, mimicking the con- hybrid genes (Fig. 2B). A common consequence of dition of FH deficiency in homozygous CFH mutation such genomic rearrangements is the formation of carriers. Thus, C3G genomic abnormalities leading to abnormal FHR proteins with duplication of SCRs aberrant FHR proteins may implicate either comple- 1 and 2. ment activation restricted to the glomeruli or fluid The two N-terminal SCRs of FHR1, FHR2, and phase complement dysregulation. Together, the earlier- FHR5 include a dimerization domain so that these described findings highlight the complexity of the FHRs always circulate in plasma as homodimers, genetics of IC-MPGN/C3G and how our knowledge heterodimers, or oligomers.64 Oligomerization increases in this area remains incomplete. 456 M.R. Noris and G. Remuzzi IGA NEPHROPATHY is consistent with an oligo/polygenic model for familial IgAN. Clinical and Histology Features and Pathogenesis GWAS in four large cohorts identified 15 distinct IgAN has been recognized as the most common form variants for IgAN overall.72,73 However, GWAS stud- of primary glomerulonephritis. The disease is most ies usually detect rather common variants, which prevalent in East Asia, accounting for up to 50% of typically show small effects, so that the overall 15 primary glomerulonephritis in biopsy registries in GWAS loci explained only 6% to 8% of the disease Japan and China, versus 10% to 35% in European risk. Nonetheless, GWAS loci highlighted some path- countries. The clinical symptoms are highly variable, ogenetic disease pathways, including antigen process- ranging from asymptomatic microhematuria to sus- ing and presentation (HLA genes), regulation of tained proteinuria and rapid deterioration of renal mucosal IgA production (LIF/OSM and TNFSF13 function. Definitive diagnosis is biopsy-based and loci), and innate immunity against intestinal pathogens defined by dominant or co-dominant glomerular depos- (DEFA, CARD9, ITGAM-ITGAX, VAV3, and a locus its of IgA.67 Electron microscopy examination showed on chromosome 1q32 including CFH and the five that the IgA deposits are localized mainly in the CFHR genes).74 mesangium. The histologic features also vary, includ- The IgAN locus on 1q32 is of great interest and ing no or minimal abnormalities by light microscopy; suggests that the alternative pathway of complement mesangial hypercellularity; focal or diffuse prolifera- plays a role in the pathogenesis of the disease. The top tive, necrotizing, and crescentic lesions; and, more SNP, c.1696þ2019G4A (rs6677604), associated rarely, membranoproliferative patterns.67 Progression with IgAN in three independent cohorts, is located to focal or diffuse segmental and global glomerulo- in intron 12 of CFH. The minor A allele of this SNP sclerosis with atrophy and interstitial fibrosis occurs in perfectly tags the CFHR3-CFHR1 deletion, and het- chronic stages. erozygous carriers were found to have a more than IgAN is an immune-mediated disease. According to 40% reduction in the risk of IgAN (Table 3).72 The the multihit model, the pathogenesis of renal injury in protection against IgAN conferred by the CFHR3– IgAN implies increased levels of IgA1 with galactose- CFHR1 deletion is the opposite of the pathogenetic deficient O-glycans (Gd-IgA1), synthesis of antibodies effects of C3G-associated CFHR duplicated/hybrid against galactose-deficient IgA1, and the formation of genes resulting in gain-of-function FHRs. In IgAN, IC that accumulate in the mesangium inducing the Gd-IgA1 containing IC deposited in the mesangium production of extracellular matrix and inflammatory could alter cell surface to favor competition between cytokines and chemokines, mesangial proliferation, FH and FHRs. The advantage of individuals carrying and activation of complement, which results in renal the CFHR3–CFHR1 deletion would be having less injury.68 FHRs that antagonize the regulatory activity of FH in the kidney mesangium. This interpretation is consis- tent with evidence of activation of the alternative Genetics pathway of complement in IgAN as shown by IF It has been postulated that genetic factors play a role in studies in kidney biopsy specimens showing consistent IgAN based on a number of findings, including differ- positive staining for C3, properdin, and C5b-9, ences in prevalence among ethnicities, familial aggre- and generally negative C1q staining.68 Moreover, gation, and disease concordance in identical twins. An increased circulating levels of activated C3 products autosomal-dominant inheritance with variable pene- and increased urinary excretion of FH, properdin, and trance is observed in most familial cases, but more C5b-9 levels have been reported in patients with IgAN complex genetic models have been identified. In than in healthy controls.75 Lectin pathway activation addition, IgA1 glycosylation defects leading to Gd- also has been suggested by glomerular staining for IgA1 have high heritability,69 although family based C4d and MBL, reported in 25% to 40% of IgAN studies showed that increased levels of Gd-IgA1 are biopsies. It is possible that glomerular deposition of not enough to induce IgAN. the C3 fragments generated by the activation of the Genome-wide linkage analyses in IgAN families lectin pathway would enhance the competition have been challenged by a high degree of genetic between FH and FHR, thus favoring the activation heterogeneity, and no clear-cut causal mutations have of the alternative pathway. In this context a lower been identified to date. A genome-wide linkage anal- copy number of CFHR1 and CHR3 genes will result ysis in 30 multiplex IgAN kindreds70 mapped the in lower alternative pathway activation and protection IgAN locus on 6q22-q23. In another genome-wide from IgAN development. scan in 22 Italian families,71 the regions 4q26-q31 and Based on some common clinical features among 17q12-q22 showed the strongest evidence of linkage. IgAN and patients with CFHR5 nephropathy, includ- Evidence of linkage to multiple chromosomal regions ing microhematuria and synpharyngitic flares with Immune-mediated glomerular diseases 457 Table 3. Complement Gene Variations Associated With IgAN and LN IgAN Gene Variant/Abnormality Main Effect Impact Frequency CFH c.1696þ2019A / Protective Common SNP (c.1696þ2019G4 A, rs 6677604) CFHR1/CFHR3 Deletion Less competition with FH Protective Common deletion CFHR5 Rare functional variants Increased competition with FH Risk 9% LN Abnormality CFH Homozygous mutations Impaired fluid phase alternative Disease-causing Very rare pathway (AP) complement regulation CFHR1/CFHR3 Deletion Anti-FH antibodies? Risk Common deletion B-cell hyperactivity? CFI Homozygous mutations Impaired fluid phase complement Disease-causing Very rare regulation C1Q, C1R, C1S Homozygous mutations C1 deficiency; lack of classical High risk of SLE and LN Rare pathway (CP) activation C4A, C4B Homozygous mutations C4 deficiency; lack of CP and High risk of SLE and LN Rare lectin pathway (LP) activation C2 Homozygous mutations C2 deficiency; lack of CP and Risk of SLE and LN Rare LP activation macrohematuria, more severe renal impairment in men, predominance of SLE with women outnumbering men and recurrence after transplantation, Zhai et al,76 by 10:1, LN affects both sexes equally. LN usually studied the CFHR5 gene in 500 patients with IgAN manifests with proteinuria, microhematuria, and and in 576 controls (Table 3). They found a similar hypertension. prevalence of CFHR5 rare variants among patients and The morphologic changes in kidney biopsies com- controls, but the location of variants along the gene and prise a spectrum of vascular, glomerular, and tubu- their effect on protein function were different in the lointerstitial lesions. Based on the degree of glomerular two groups. In particular, three IgAN-associated rare involvement, LN currently is classified histologically variants resulted in FHR5 proteins with enhanced into the following: class I, with mesangial immune binding to C3b in vitro. The investigators hypothesized deposits without mesangial hypercellularity; class II, that these variants generate gain-of-function FHR5 with mesangial immune deposits with mesangial molecules with stronger competitive activity with FH, hypercellularity; class III for focal proliferative glo- thus contributing to complement activation and IgAN merulonephritis; class IV with diffuse glomeruloneph- susceptibility. However, one IgAN-associated variant ritis; class V for membranous lupus nephritis; and resulted in a truncated protein with lower C3b binding class VI for advanced sclerosing lesions.78 Advanced- capacity than wild-type protein. In addition, the stage sclerosing class III or class IV LN and inactive P value for the distribution of rare variants is only sclerosing class VI are associated clinically with more suggestive, and differences in ethnicity among patients severe chronic kidney disease and the risk of progres- and controls may confound the results because sion to end-stage renal disease. rare variant tests are sensitive even to subtle Autoantibodies are crucial to the pathogenesis of ancestry differences. Thus, more studies are needed both SLE and LN. Commonly, autoantibodies are before unequivocally declaring CFHR5 a new IgAN directed against nucleic acids and proteins involved susceptibility gene. in RNA transcription and protein translation, such as double-stranded DNA, Sm antigen, and nucleosomes. During early stages of the disease there is defective LUPUS NEPHRITIS clearance of apoptotic cells, resulting in abnormal exposure of autoantigens through nuclear proteins Clinical and Histology Features and Pathogenesis clustering on the blebs of apoptotic cells.79 Polyclonal LN is an immune complex–mediated glomeruloneph- hyperactivity of B cells, germ-line mutations leading to ritis that is a severe feature of systemic lupus eryth- clonal expansion of autoreactive B cells, and increased ematosus (SLE), occurring in approximately 40% of production of B-cell–stimulating cytokines are thought patients, which contributes substantially to SLE-related to underlie autoantibody production. T-cell hyperac- morbidity and mortality.77 Although there is a gender tivity and defective T-cell tolerance has been proposed 458 M.R. Noris and G. Remuzzi to contribute to B-cell proliferation and to the produc- the G4T STAT4 variant (rs7574865, in intron 3) that tion of autoantibodies. The result of the earlier- is associated with increased STAT4 RNA levels, was described events is the production of IC, which, if found to be over-represented in SLE patients in differ- not cleared adequately, may deposit in the kidney. ent populations,83 and the association was more strik- Immune complex localization in the glomerulus leads ing in patients with nephritis.83,84 The role of genetic to an inflammatory response with leukocyte infiltration, variations affecting the type I interferon activity in LN the activation of Fc receptors, and the release of pathogenesis has been confirmed by association studies proteolytic enzymes and cytokines associated with on the interferon regulatory factor (IRF) genes, encod- cellular proliferation and matrix formation, whereas ing interferon regulatory factors that regulate the effect the binding of nucleosome to Toll-like receptor 2 and of interferon on signaling and immune cell develop- Toll-like receptor 9 activates renal resident dendritic ment. Thus, a higher frequency of the SLE risk T allele cells. The complement system plays a dual role in SLE of the IRF5 intronic SNP, c.-12þ198G4T, which and LN. On the one hand complement components are alters IRF5 splicing, allowing the expression of several useful, aiding in the apoptotic clearance and removal of unique messenger RNA isoforms, was found in more immune complexes. However, the IC activate the LN patients than in healthy subjects in a Chinese complement classic pathway in the glomeruli, as population.85 Another study84 identified an association documented by glomerular positivity for C1q, C4d, in Caucasian patients between SLE and LN and the and C3 staining, and may contribute to amplifying the risk C allele of the c.*128T4C SNP, in the 3’ inflammatory response. untranslated region of the gene, which alters the stability of IRF5 messenger RNAs. GENETICS The contribution of genetic factors to SLE and LN is Complement Genetic Variations supported by the clustering in families, the racial Convincing evidence regarding the role of genetic differences in susceptibility (both SLE and LN are variations affecting the complement system (Table 3) three- to four-fold more common in blacks, Asians, derives from studies in rare autosomal-recessive forms and Hispanics than in Caucasians), and concordance of of complement component deficiencies. Homozygous SLE in more than 20% of identical twins. or compound heterozygous mutations leading to com- Variants in more than 30 genes that appear to be plete deficiency of each of the classic pathway compo- associated with SLE have been identified, including nents (C1q, C1r, C1s, C4, and C2) are associated with HLA genes, genes encoding IC receptors, cytokines high susceptibility to SLE. C1q deficiency is rare, with and chemokines and their signaling molecules, and fewer than 50 cases reported in the literature.86 components of the innate immune system.79 Several of Approximately 90% of these patients had SLE or a these gene variants impart susceptibility to LN. lupus-like syndrome, which often was severe. Of the The HLA DR3 allele (DRB1*0301) is a SLE risk affected patients, more than 70% had high titers of allele, and in SLE patients it is associated with the risk autoantibodies, mainly antinuclear antibodies and, of developing LN,80 proliferative nephritis, and with more rarely, anti–double-stranded DNA antibodies, the production of anti–double-stranded DNA antibod- and 42% developed glomerulonephritis.87 Hereditary ies,80 suggesting a genetic contribution to the produc- deficiency of C1s and C1r is even more rare than tion of antibodies targeting the kidney. The R variant C1q deficiency (o20 cases reported),86 and most of the p.H131R polymorphisms in FcγRIIA has been cases have deficiencies of both components. SLE associated with SLE and LN across multiple ethnic- occurs in approximately 50% of these individuals,87 ities.81 This variant has a lower binding affinity for whereas nephritis was reported only occasionally IgG2, thereby reducing the clearance of IC, which (Table 3). contributes to LN pathogenesis. Similarly, a large C4 protein is encoded by two genes (C4A and meta-analysis comparison of SLE patients with or C4B), which share 99% sequence identity and are without LN showed an over-representation of the low located in the major histocompatibility complex IgG1 and IgG3-binding F allele of the p.V176F (MHC) class III cluster in chromosome 6p21.3. The FcγRIIIA SNP, among patients who developed renal C4 gene copy number ranges from two to eight; the disease.82 more common copy number in healthy populations is Several studies have identified an association two of each isotype. C4 deficiency can arise from copy between SLE and genetic variants in the signal trans- number variation at the C4 locus, as well as from ducer and activator of transcription (STAT)4 gene mutations in the gene sequence, resulting in nonex- encoding a transcription factor that transmits signals pressed (or null) C4 alleles (called C4AQ*0 and by several cytokines, including interleukin-12, inter- C4BQ*0). The most frequent mutation is a two–base leukin-23, and type I interferons. The minor T allele of pair cytosine thymidine insertion in exon 29, which Immune-mediated glomerular diseases 459 causes a frameshift change and a premature stop codon B-cell function. On the one hand, they promote the in exon 30.88 production of antibodies by favoring the maturation of Approximately 30 cases of complete C4 deficiency B cells and reducing the threshold for B-cell activa- have been reported. This deficiency is related strongly tion.4 In this view, complement deficiency would be to SLE, with a prevalence of 75%.89 Approximately predicted to impair antibody response and protect from 50% of SLE patients with C4 deficiency develop autoimmunity associated with SLE. Conversely, inter- glomerulonephritis and more than 70% have anti- action between C4 activation fragments and CR1 nuclear antibodies (ANA) and anti-Rho autoantibodies regulates B-cell tolerance by preventing autoreactive (Table 3).90 B cells from maturing into plasma cells,94 and C1q Partial C4 deficiency caused by C4AQ*0 alleles also inhibits T-cell proliferation through gC1q/C1q inter- has been associated with SLE. Relative risk ranges actions, a function that is defective in patients with between 10 and 16 for homozygous C4AQ*0 carriers, SLE.94 The association between low CR1 expression to 2 to 5 in heterozygous carriers.91 However, it has and gene transcription with disease severity and renal been difficult to establish whether this association is involvement in patients with SLE95 supports the related primarily to C4AQ*0 or secondary to long- hypothesis that defective C4/CR1 signaling plays a range linkage disequilibrium across the MHC region. role in the pathogenesis of LN. Indeed, the major source of C4AQ*0 alleles is the Finally, very rare patients with inherited homozy- common European MHC haplotype: HLA-A1, HLA- gous mutations in CFH96 or CFI97 who developed SLE B8, HLA-DR3, which shows an association with SLE. with glomerulonephritis (Table 3) would indicate that A recent large study in more than 1,000 SLE patients uncontrolled activation of the alternative B cell recep- and controls found that partial C4 deficiency is not an tor pathway plays a role in LN. Because FH-deficient independent risk factor for SLE.92 Univariate analysis and FI-deficient patients cannot regulate the alternative documented an association with SLE and an intronic pathway properly, these patients also have low levels SNP in the C2 gene (rs558702), the C4A zero copy of FB and properdin, reflecting constant activation of allele, and HLA-DR3. However, in multivariate logis- this pathway. Low levels of C5 also indicate massive tic regression, conditioning for rs558702 or HLA-DR3 activation of the terminal complement pathway. The abrogated the effect of the C4A zero copy allele, terminal complement products C5a and C5b-9, indicating that the association of SLE with this allele together with activation of C3 in the circulation and is owing mainly to its linkage with polymorphisms in locally produced by infiltrating leukocytes and glomer- the MHC region.92 ular cells, could trigger an inflammatory response and Homozygous C2 deficiency has a 1:10,000 to tissue injury in the kidney. Interestingly, a large 1:20,000 prevalence in Caucasians.86 The majority of association study in four ethnic groups showed that homozygous carriers are asymptomatic, only approxi- CFHR3–CFHR1 deletion is a predisposing factor for mately 10% to 30% develop SLE. The severity of SLE SLE (Table 3).98 Whether this association could be in C2 deficiency is similar to that of SLE in other explained by the capacity of FHR1 to inhibit alter- patients. Severe kidney involvement appears to be rare, native pathway C5 convertase activity and the forma- but may occur. Homozygous C3 deficiency, reported in tion of C5b-9,64 or by the possibility that carriers of the fewer than 30 subjects,86 predisposes to recurrent bacte- deletion form autoantibodies against factor H, similar rial infections in childhood and MPGN, but SLE is rare. to the situation with aHUS, has not been addressed yet. Together, these findings suggest that the comple- Another hypothesis could be that FHR3 deficiency ment classic pathway plays a role in protecting against may predispose to B-cell hyperactivity, as suggested the development of SLE and its renal complications. In by very recent findings that FHR3 binding to C3d and the pathogenesis of LN, several factors are relevant, C3b fragments on B cells blocked B-cell activation including the role of complement in the clearance and through the B cell receptor co-receptor complex.99 processing of IC, through binding erythrocyte CR1 to C3b/C4b–coated IC, and the removal of apoptotic cells CONCLUSIONS and debris, through the binding of C1q, C4b, and C3b, which enhances the ingestion of dead cells by phag- The identification of mutations and genomic rearrange- ocytes.93 Thus, complement deficiency, as a conse- ments in complement genes, and functional studies on quence of defective IC processing, may favor IC mutants, clearly have shown the role of alternative deposition in the kidneys, and predispose to SLE and pathway dysregulation in aHUS, and genotyping has its renal complication. Similarly, defective clearance of entered clinical practice. These studies have paved the apoptotic cells would result in increased autoantigen way for clinical trials with the anti-C5 antibody load, driving the autoimmune response of SLE. eculizumab, which has become the standard of care Complement products and their interaction with the in aHUS. Evidence also is accumulating regarding the receptors CR1 and CR2 on B cells have a dual role in association between complement gene abnormalities 460 M.R. Noris and G. Remuzzi and IC-MPGN and C3G. Although these conditions and heparin and surface attachment to endothelial cells in are rare, their study has provided important insights hemolytic uremic syndrome. J Clin Invest. 2003;111:1181-90. 12. Venables JP, Strain L, Routledge D, Bourn D, Powell HM, into the pathogenesis of complement-mediated renal Warwicker P, et al. Atypical haemolytic uraemic syndrome tissue injury. Complement genetic studies in more associated with a hybrid complement gene. PLoS Med. 2006;3: common and complex glomerular diseases, such as e431. IgAN and LN, have broadened our understanding of 13. Challis RC, Araujo GS, Wong EK, Anderson HE, Awan A, the fragile balance between the protective and harmful Dorman AM, et al. A de novo deletion in the regulators of complement activation cluster producing a hybrid complement functions of the complement system. The activation of factor H/complement factor H-related 3 gene in atypical the early steps of the classic complement pathways hemolytic uremic syndrome. J Am Soc Nephrol. 2016;27: likely is instrumental for limiting IC load in the 1617-24. circulation and removing IC and damaged cells from 14. Valoti E, Alberti M, Tortajada A, Garcia-Fernandez J, Gastoldi the glomerulus, whereas the dysregulation of the S, Besso L, et al. A novel atypical hemolytic uremic syndrome- associated hybrid CFHR1/CFH gene encoding a fusion protein alternative pathway and the activation of the common that antagonizes factor H-dependent complement regulation. terminal pathway mediate tissue injury. These findings J Am Soc Nephrol. 2015;26:209-19. may be relevant for the development of new targeted 15. Dragon-Durey MA, Sinha A, Tagarsimalemath SK, Bagga A. complement inhibitors for immune-mediated glomer- Anti-complement-factor H-associated glomerulopathies. Nat ular diseases. Rev Nephrol. 2016;12:563-78. 16. Jozsi M, Licht C, Strobel S, Zipfel SL, Richter H, Heinen S, et al. Factor H autoantibodies in atypical hemolytic uremic syndrome correlate with CFHR1/CFHR3 deficiency. Blood. ACKNOWLEDGMENTS 2008;111:1512-4. The authors thank Kerstin Mierke for English editing 17. Noris M, Brioschi S, Caprioli J, Todeschini M, Bresin E, and Manuela Passera for secretarial assistance. Porrati F, et al. Familial haemolytic uraemic syndrome and an MCP mutation. Lancet. 2003;362:1542-7. 18. Richards A, Kemp EJ, Liszewski MK, Goodship JA, Lampe AK, Decorte R, et al. Mutations in human complement REFERENCES regulator, membrane cofactor protein (CD46), predispose to 1. Dickinson BL. Unraveling the immunopathogenesis of glomer- development of familial hemolytic uremic syndrome. Proc Natl ular disease. Clin Immunol. 2016;169:89-97. Acad Sci U S A. 2003;100:12966-71. 2. Noris M, Remuzzi G. Atypical hemolytic-uremic syndrome. 19. Kavanagh D, Kemp EJ, Mayland E, Winney RJ, Duffield JS, N Engl J Med. 2009;361:1676-87. Warwick G, et al. Mutations in complement factor I predispose 3. Noris M, Remuzzi G. Glomerular diseases dependent on to development of atypical hemolytic uremic syndrome. J Am complement activation, including atypical hemolytic uremic Soc Nephrol. 2005;16:2150-5. syndrome, membranoproliferative glomerulonephritis, and C3 20. Bienaime F, Dragon-Durey MA, Regnier CH, Nilsson SC, glomerulopathy: core curriculum 2015. Am J Kidney Dis. Kwan WH, Blouin J, et al. Mutations in components of 2015;66:359-75. complement influence the outcome of Factor I-associated 4. Noris M, Remuzzi G. Overview of complement activation and atypical hemolytic uremic syndrome. Kidney Int. 2010;77: regulation. Semin Nephrol. 2013;33:479-92. 339-49. 5. Noris M, Galbusera M, Gastoldi S, Macor P, Banterla F, Bresin 21. Fremeaux-Bacchi V, Miller EC, Liszewski MK, Strain L, E, et al. Dynamics of complement activation in aHUS and how Blouin J, Brown AL, et al. Mutations in complement C3 to monitor eculizumab therapy. Blood. 2014;124:1715-26. predispose to development of atypical hemolytic uremic syn- 6. Noris M, Ruggenenti P, Perna A, Orisio S, Caprioli J, Skerka drome. Blood. 2008;112:4948-52. C, et al. Hypocomplementemia discloses genetic predisposition 22. Goicoechea de Jorge E, Harris CL, Esparza-Gordillo J, Carreras to hemolytic uremic syndrome and thrombotic thrombocyto- L, Arranz EA, Garrido CA, et al. Gain-of-function mutations in penic purpura: role of factor H abnormalities. Italian Registry of complement factor B are associated with atypical hemolytic Familial and Recurrent Hemolytic Uremic Syndrome/Throm- uremic syndrome. Proc Natl Acad Sci U S A. 2007;104:240-5. botic Thrombocytopenic Purpura. J Am Soc Nephrol. 1999; 23. Schramm EC, Roumenina LT, Rybkine T, Chauvet S, Vieira- 10:281-93. Martins P, Hue C, et al. Mapping interactions between comple- 7. Kerr H, Richards A. Complement-mediated injury and protec- ment C3 and regulators using mutations in atypical hemolytic tion of endothelium: lessons from atypical haemolytic uraemic uremic syndrome. Blood. 2015;125:2359-69. syndrome. Immunobiology. 2012;217:195-203. 24. Marinozzi MC, Vergoz L, Rybkine T, Ngo S, Bettoni S, 8. Legendre CM, Licht C, Muus P, Greenbaum LA, Babu S, Pashov A, et al. Complement factor B mutations in atypical Bedrosian C, et al. Terminal complement inhibitor eculizumab hemolytic uremic syndrome–disease-relevant or benign?. J Am in atypical hemolytic-uremic syndrome. N Engl J Med. 2013; Soc Nephrol. 2014;25:2053-65. 368:2169-81. 25. Delvaeye M, Noris M, De Vriese A, Esmon CT, Esmon NL, 9. Noris M, Caprioli J, Bresin E, Mossali C, Pianetti G, Gamba S, Ferrell G, et al. Thrombomodulin mutations in atypical et al. Relative role of genetic complement abnormalities in hemolytic-uremic syndrome. N Engl J Med. 2009;361:345-57. sporadic and familial aHUS and their impact on clinical 26. Lhotta K, Janecke AR, Scheiring J, Petzlberger B, Giner T, phenotype. Clin J Am Soc Nephrol. 2010;5:1844-59. Fally V, et al. A large family with a gain-of-function mutation 10. Warwicker P, Goodship TH, Donne RL, Pirson Y, Nicholls A, of complement C3 predisposing to atypical hemolytic uremic Ward RM, et al. Genetic studies into inherited and sporadic syndrome, microhematuria, hypertension and chronic renal hemolytic uremic syndrome. Kidney Int. 1998;53:836-44. failure. Clin J Am Soc Nephrol. 2009;4:1356-62. 11. Manuelian T, Hellwage J, Meri S, Caprioli J, Noris M, Heinen 27. Martinez-Barricarte R, Pianetti G, Gautard R, Misselwitz J, S, et al. Mutations in factor H reduce binding affinity to C3b Strain L, Fremeaux-Bacchi V, et al. The complement factor H Immune-mediated glomerular diseases 461 R1210C mutation is associated with atypical hemolytic uremic 43. Hou J, Markowitz GS, Bomback AS, Appel GB, Herlitz LC, syndrome. J Am Soc Nephrol. 2008;19:639-46. Barry Stokes M, et al. Toward a working definition of C3 28. Bresin E, Rurali E, Caprioli J, Sanchez-Corral P, Fremeaux- glomerulopathy by immunofluorescence. Kidney Int. 2014;85: Bacchi V, Rodriguez de Cordoba S, et al. Combined comple- 450-6. ment gene mutations in atypical hemolytic uremic syndrome 44. Iatropoulos P, Noris M, Mele C, Piras R, Valoti E, Bresin E, influence clinical phenotype. J Am Soc Nephrol. 2013;24: et al. Complement gene variants determine the risk of 475-86. immunoglobulin-associated MPGN and C3 glomerulopathy 29. Esparza-Gordillo J, Jorge EG, Garrido CA, Carreras L, and predict long-term renal outcome. Mol Immunol. 2016;71: Lopez-Trascasa M, Sanchez-Corral P, et al. Insights into 131-42. hemolytic uremic syndrome: segregation of three independent 45. Alfandary H, Davidovits M. Novel factor H mutation associ- predisposition factors in a large, multiple affected pedigree. ated with familial membranoproliferative glomerulonephritis Mol Immunol. 2006;43:1769-75. type I. Pediatr Nephrol. 2015;30:2129-34. 30. Caprioli J, Castelletti F, Bucchioni S, Bettinaglio P, Bresin E, 46. Dragon-Durey MA, Fremeaux-Bacchi V, Loirat C, Blouin J, Pianetti G, et al. Complement factor H mutations and gene Niaudet P, Deschenes G, et al. Heterozygous and homozygous polymorphisms in haemolytic uraemic syndrome: the C-257T, factor H deficiencies associated with hemolytic uremic syn- the A2089G and the G2881T polymorphisms are strongly drome or membranoproliferative glomerulonephritis: report and associated with the disease. Hum Mol Genet. 2003;12:3385-95. genetic analysis of 16 cases. J Am Soc Nephrol. 2004;15: 31. Pickering MC, de Jorge EG, Martinez-Barricarte R, Recalde S, 787-95. Garcia-Layana A, Rose KL, et al. Spontaneous hemolytic 47. Licht C, Heinen S, Jozsi M, Loschmann I, Saunders RE, uremic syndrome triggered by complement factor H lacking Perkins SJ, et al. Deletion of Lys224 in regulatory domain surface recognition domains. J Exp Med. 2007;204:1249-56. 4 of factor H reveals a novel pathomechanism for dense deposit 32. Tortajada A, Montes T, Martinez-Barricarte R, Morgan BP, disease (MPGN II). Kidney Int. 2006;70:42-50. Harris CL, de Cordoba SR. The disease-protective complement 48. Servais A, Noel LH, Dragon-Durey MA, Gubler MC, Remy P, factor H allotypic variant Ile62 shows increased binding affinity Buob D, et al. Heterogeneous pattern of renal disease associated for C3b and enhanced cofactor activity. Hum Mol Genet. with homozygous factor H deficiency. Hum Pathol. 2011;42: 2009;18:3452-61. 1305-11. 33. Esparza-Gordillo J, Goicoechea de Jorge E, Buil A, Carreras 49. Zhang Y, Meyer NC, Wang K, Nishimura C, Frees K, Jones M, Berges L, Lopez-Trascasa M, Sanchez-Corral P, et al. Predis- et al. Causes of alternative pathway dysregulation in dense position to atypical hemolytic uremic syndrome involves the deposit disease. Clin J Am Soc Nephrol. 2012;7:265-74. concurrence of different susceptibility alleles in the regulators 50. Prasto J, Kaplan BS, Russo P, Chan E, Smith RJ, Meyers KE. of complement activation gene cluster in 1q32. Hum Mol Streptococcal infection as possible trigger for dense deposit Genet. 2005;14:703-12. disease (C3 glomerulopathy). Eur J Pediatr. 2014;173:767-72. 34. Abarrategui-Garrido C, Martinez-Barricarte R, Lopez-Trascasa 51. Bu F, Borsa NG, Jones MB, Takanami E, Nishimura C, Hauer M, de Cordoba SR, Sanchez-Corral P. Characterization of JJ, et al. High-throughput genetic testing for thrombotic micro- complement factor H-related (CFHR) proteins in plasma angiopathies and C3 glomerulopathies. J Am Soc Nephrol. reveals novel genetic variations of CFHR1 associated with 2016;27:1245-53. atypical hemolytic uremic syndrome. Blood. 2009;114: 52. Recalde S, Tortajada A, Subias M, Anter J, Blasco M, Maranta 4261-71. R, et al. Molecular basis of factor H R1210C association with 35. Mele C, Lemaire M, Iatropoulos P, Piras R, Bresin E, Bettoni S, ocular and renal diseases. J Am Soc Nephrol. 2016;27:1305-11. et al. Characterization of a new DGKE intronic mutation in 53. Martinez-Barricarte R, Heurich M, Valdes-Canedo F, Vazquez- genetically unsolved cases of familial atypical hemolytic Martul E, Torreira E, Montes T, et al. Human C3 mutation uremic syndrome. Clin J Am Soc Nephrol. 2015;10:1011-9. reveals a mechanism of dense deposit disease pathogenesis and 36. Lemaire M, Fremeaux-Bacchi V, Schaefer F, Choi M, Tang provides insights into complement activation and regulation. WH, Le Quintrec M, et al. Recessive mutations in DGKE cause J Clin Invest. 2010;120:3702-12. atypical hemolytic-uremic syndrome. Nat Genet. 2013;45: 54. Chauvet S, Roumenina LT, Bruneau S, Marinozzi MC, 531-6. Rybkine T, Schramm EC, et al. A familial C3GN secondary 37. Bruneau S, Neel M, Roumenina LT, Frimat M, Laurent L, to defective C3 regulation by complement receptor 1 and factor Fremeaux-Bacchi V, et al. Loss of DGKepsilon induces H. J Am Soc Nephrol. 2016;27:1665-77. endothelial cell activation and death independently of comple- 55. Abrera-Abeleda MA, Nishimura C, Frees K, Jones M, Maga T, ment activation. Blood. 2015;125:1038-46. Katz LM, et al. Allelic variants of complement genes associated 38. Sanchez Chinchilla D, Pinto S, Hoppe B, Adragna M, Lopez L, with dense deposit disease. J Am Soc Nephrol. 2011;22: Justa Roldan ML, et al. Complement mutations in diacylgly- 1551-9. cerol kinase-epsilon-associated atypical hemolytic uremic syn- 56. Herbert AP, Deakin JA, Schmidt CQ, Blaum BS, Egan C, drome. Clin J Am Soc Nephrol. 2014;9:1611-9. Ferreira VP, et al. Structure shows that a glycosaminoglycan 39. Sethi S, Fervenza FC. Membranoproliferative glomeruloneph- and protein recognition site in factor H is perturbed by age- ritis–a new look at an old entity. N Engl J Med. 2012;366: related macular degeneration-linked single nucleotide polymor- 1119-31. phism. J Biol Chem. 2007;282:18960-8. 40. Pickering MC, D’Agati VD, Nester CM, Smith RJ, Haas M, 57. Conway EM. Thrombomodulin and its role in inflammation. Appel GB, et al. C3 glomerulopathy: consensus report. Kidney Semin Immunopathol. 2012;34:107-25. Int. 2013;84:1079-89. 58. Wilson V, Darlay R, Wong W, Wood KM, McFarlane J, 41. Servais A, Noel LH, Roumenina LT, Le Quintrec M, Ngo S, Schejbel L, et al. Genotype/phenotype correlations in comple- Dragon-Durey MA, et al. Acquired and genetic complement ment factor H deficiency arising from uniparental isodisomy. abnormalities play a critical role in dense deposit disease and Am J Kidney Dis. 2013;62:978-83. other C3 glomerulopathies. Kidney Int. 2012;82:454-64. 59. Gale DP, de Jorge EG, Cook HT, Martinez-Barricarte R, 42. Sethi S, Nester CM, Smith RJ. Membranoproliferative glomer- Hadjisavvas A, McLean AG, et al. Identification of a mutation ulonephritis and C3 glomerulopathy: resolving the confusion. in complement factor H-related protein 5 in patients of Cypriot Kidney Int. 2012;81:434-41. origin with glomerulonephritis. Lancet. 2010;376:794-801. 462 M.R. Noris and G. Remuzzi 60. Tortajada A, Yebenes H, Abarrategui-Garrido C, Anter J, systemic lupus erythematosus revisited. Kidney Int. 2004;65: Garcia-Fernandez JM, Martinez-Barricarte R, et al. C3 521-30. glomerulopathy-associated CFHR1 mutation alters FHR oligo- 79.