sign in sign up
<<
Home , CFHR3, CFHR4, CFHR5, Factor H

CLINICAL RESEARCH www.jasn.org

Rare Variants in the Complement –Related 5 Contribute to Genetic Susceptibility to IgA Nephropathy

†‡ †‡ †‡ †‡ †‡ Ya-Ling Zhai,* § Si-Jun Meng,* § Li Zhu,* § Su-Fang Shi,* § Su-Xia Wang,* § †‡ †‡ †‡ †‡ †‡ Li-Jun Liu,* § Ji-Cheng Lv,* § Feng Yu,* § Ming-Hui Zhao,* § and Hong Zhang* §

*Renal Division, Department of Medicine, Peking University First Hospital, Beijing, China; †Institute of Nephrology, Peking University, Beijing, China; ‡Key Laboratory of Renal Disease, Ministry of Health of China, Beijing, China; and §Key Laboratory of Chronic Kidney Disease Prevention and Treatment (Peking University), Ministry of Education, Beijing, China

ABSTRACT A recent genome–wide association study of IgA nephropathy (IgAN) identified 1q32, which contains multiple complement regulatory , including the complement factor H (CFH) gene and the comple- ment factor H–related (CFHRs) genes, as an IgAN susceptibility . Abnormal complement activation caused by a in CFHR5 was shown to cause CFHR5 nephropathy, which shares many character- istics with IgAN. To explore the genetic effect of variants in CFHR5 on IgAN susceptibility, we recruited 500 patients with IgAN and 576 healthy controls for genetic analysis. We sequenced all exons and their intronic flanking regions as well as the untranslated regions of CFHR5 and compared the frequencies of identified variants using the sequence kernel association test. We identified 32 variants in CFHR5, includ- ing 28 rare and four common variants. The distribution of rare variants in CFHR5 in patients with IgAN differed significantly from that in controls (P=0.002). Among the rare variants, in silico programs predicted nine as potential functional variants, which we then assessed in functional assays. Compared with wild-type CFHR5, three recombinant CFHR5 , CFHR5-M (c.508G.A/p.Val170Met), CFHR5-S (c.533A.G/p. Asn178Ser), and CFHR5-D (c.822A.T/p.Glu274Asp), showed significantly higher binding capacity (CFHR5-M: 109.67%63.54%; P=0.02; CFHR5-S: 174.27%69.78%; P,0.001; CFHR5-D: 127.25%61.75%; P,0.001), whereas another recombinant CFHR5 (c.776T.A/p.Leu259Termination) showed less C3b bind- ing (56.89%60.57%; P,0.001). Our study found that rare variants in CFHR5 may contribute to the genetic susceptibility to IgAN, which suggests that CFHR5 is an IgAN susceptibility gene.

J Am Soc Nephrol 27: 2894–2905, 2016. doi: 10.1681/ASN.2015010012

IgA nephropathy (IgAN) is the most common pri- IgAN. A recent genome–wide association study mary GN worldwide.1 Recently, important discov- (GWAS)7 and a GWAS replication study8 identified eries have been made regarding the pathogenesis of and confirmed 1q32 as an IgAN susceptibility IgAN,2–5 such as the aberrant glycosylation of IgA1 and the presence of antiglycan in circu- lation as well as the role of receptors in Received January 5, 2015. Accepted December 20, 2015. glomeruli. However, there are still many unknown Y.-L.Z. and S.-J.M. contributed equally to this work. aspects in IgAN. Although complement activation Published online ahead of print. Publication date available at has been widely accepted for a long time to occur in www.jasn.org. IgAN,6 the exact pathways and regulatory mecha- Correspondence: Dr. Li Zhu, Renal Division, Department of nisms are still unclear. Medicine, Peking University First Hospital, Peking University In- Meanwhile, from racial and ethnic variations in stitute of Nephrology, Beijing 100034, China. Email: funnyzhuli@ disease incidence to familial aggregation of IgAN, bjmu.edu.cn increasing evidence implied a genetic effect on Copyright © 2016 by the American Society of Nephrology

2894 ISSN : 1046-6673/2709-2894 JAmSocNephrol27: 2894–2905, 2016 www.jasn.org CLINICAL RESEARCH locus, which suggested the involvement of genetic abnormal- Their proteinuria and eGFR levels were 1.42 (0.81–2.66) g/d ities in the complement factor H (CFH)and/orthecomplement and 84.19627.03 ml/min per 1.73 m2, respectively. In total, factor H–related (CFHRs) genes in IgAN. This region contains 13.60% (68 of 500) had hypertension. Regarding pathologic several complement regulatory genes, CFH-CFHRs. CFHRs, features, 47.60% of patients with IgAN had a strong intensity which include CFHR3, CFHR1, CFHR4, CFHR2,andCFHR5, of C3 deposition (3+ or 4+). Haas 1–5 were found in 8.60%, share high with CFH. Accordingly, they 0.80%, 45.20%, 33.20%, and 12.20% of patients, respectively. have similar structures, which are built on a motif of distinct All patients were regularly followed for at least 1 year, with a functional domains called short consensus repeats (SCRs), mean follow-up time of 56.26 months. Patients’ treatment and and similar functions in complement regulation,9 although follow-up were conducted at an IgAN-specific clinic in Peking their precise biologic roles are not completely identical. University First Hospital and consistent with the Kidney Recently, a novel disease named CFHR5 nephropathy was Disease Improving Global Outcomes guideline.14 During identified after genetic analysis of a pedigree of Cypriot ances- follow-up, 16.80% (84 patients) of these patients reached a try with primary GN.10 Patients with CFHR5 nephropathy composited end point (ESRD, 50% eGFR decline, or death). presented with many characteristics, including young age The clinical, laboratory, and pathologic characteristics of the onset, tendency in men, repeated episodes of synpharyngitic recruited patients were similar to those of previously reported hematuria, mesangial matrix expansion, increased glomerular IgAN cohorts.15 cellularity, glomerular staining for C3, mesangial electron– dense deposits, and high recurrence rate after renal transplan- Identification of Genetic Variants in CFHR5 tation, which shared many similarities with IgAN.11 Patients All exons in the CFHR5 gene were amplified and screened for with CFHR5 nephropathy bear the same heterozygous inter- genetic variants in recruited individuals. In total, 32 variants nal duplication of the CFHR5 gene, which is likely to be the were identified (Figure 1, Table 1), including 10 in exons, 18 in cause of their glomerular lesions. introns, two in promoter, and two in 39 untranslated region The CFHR5 gene is located at 1q31.3 and spans 42 kb of the (UTR). Of 10 exonic variants, eight were nonsynonymous . Its encoded product is a 62-kD protein coding (including one nonsynonymous stop gain and nonsy- named CFHR5. The CFHR5 gene has 10 exons encoded in nonymous stop loss), and two were synonymous changes. nine SCRs, in which SCR3–7andSCR8–9 showed a high ho- Among 32 variants, 16 were found in the dbSNP database, mology with SCR10–14 and SCR19–20 in CFH, respectively. whereas the other 16 variants were novel. CFHR5 is the most recently discovered CFHR protein, and it is also the one that showed the most similarity with CFH among Genetic Association Analyses of CFHR5 with IgAN the five CFHRs.12 As with CFH and other CFHRs, CFHR5 is Among the identified 32 variants, four were common variants produced in the liver, circulates in plasma, and functions as a (c.58+17T.A, c.431–80G.A, c.434G.A, and c.1147+24T. complement regulatory protein. Until today, the precise G; minor allele frequency [MAF] .0.01 in controls), whereas mechanism for CFHR5 in complement regulation is still un- the other 28 were rare. We only observed individuals with known.13 homozygous minor alleles in two common variants (c.58+ However, with respect to IgAN, the disease presents 17T.A and c.1147+24T.G) and not in the other two com- complement activation in glomeruli and has clinical features mon variants (c.431–80G.A [MAF=0.013] and c.434G.A similar to CFHR5 nephropathy, but whether some patients [MAF=0.013]) and 28 rare variants. Of 28 rare variants, 14 with IgAN have genetic changes in the CFHR5 gene and were identified only in patients with IgAN (who exhibited a whether these variants contribute to IgAN susceptibility are total of 22 rare variants), nine were identified only in healthy still unknown. In this study, using a large population of pa- controls (with a total of 23 rare variants), and the other five tients with IgAN, we performed an intensive genetic analysis variants were identified in both patients (23 total rare variants) to explore the genetic effect of variants in the CFHR5 gene on and controls (21 total rare variants). IgAN susceptibility. Because analysis of rare variants requires statistical methods that are fundamentally different than testing commonvariants, we applied Sequence Kernel Association Test (SKAT), which RESULTS can evaluate not only rare variants’ effect but also, the cumu- lative effect of rare and common variants. The results of SKAT Baseline Clinical, Laboratory, and Pathologic showed that the combined effect of rare and common variants Characteristics in CFHR5 had a significant association with IgAN (P=0.03) In total, 1076 individuals were recruited, including 500 patients (Table 2). Furthermore, we investigated common and rare with IgAN (258 men and 242 women) and 576 healthy controls variants separately. Rare variants in CFHR5 presented a sig- (310 men and 266 women). The average ages for patients with nificant association with IgAN (P=0.002) (Table 2), but com- IgAN and controls were 34.23611.09 and 33.5067.66 years mon variants did not (P=0.92) (Table 2). old, respectively. For patients with IgAN, detailed clinical, lab- Regarding the four common variants, we also performed oratory, and pathologic characteristics were also available. single variant–level analysis. In accordance with the results

J Am Soc Nephrol 27: 2894–2905, 2016 CFHR5 in IgAN 2895 CLINICAL RESEARCH www.jasn.org

c.1513+126A.G in three patients; c.254– 66A.T in two patients; and the following 11 variants, c.-350G.T, c.332A.C, c.430+ 112T.C, c.776T.A, c.790+2T.C, c.790+47A.G, c.1148–121G.A, c.1330+ 78A.G, c.1708T.C, c.1710*3C.G, and c.1710*168G.T, in one patient each). According to these rare variants in CFHR5, patients with IgAN were divided into two groups: those with rare variants in CFHR5 (43 patients) and those without (457 patients). However, clinical and path- ologic manifestations showed no signifi- cant differences between the two groups (Table 4). For long–term renal survival, no significant difference was observed be- tween patients in the two groups (Figure 2).

In Silico Functional Prediction of Rare Variants in CFHR5 Because rare variants in CFHR5 showed significant association with IgAN, we next explored their effect on CFHR5 expression, structure, and function. Several types of software were used according to the posi- tion of the variants. In total, nine of 28 rare variants were predicted to influence func- tion or expression of CFHR5 (Figure 3, Table5).Ofthesenine,five variants Figure 1. Flowchart of the genetic analysis of CFHR5 variants in IgAN. After screening might result in potential pathogenic amino for variants in CFHR5 in 500 patients with IgAN and 576 healthy controls, 32 variants . were identified. After checking against the dbSNP database, 1000 Genome data, and acid changes, including c.332A C/p. fi ExAC data to see whether these variants were reported before, 32 variants in CFHR5 Gln111Pro (identi ed in patients and con- were divided into two groups, common variants and rare variants, on the basis of the trols) in exon 3, c.508G.A/p.Val170Met MAF in controls (cutoff =0.01). Then, variant distribution between patients with IgAN (identified in patients and controls) and and controls was compared in each group using the SKAT. After the identification of c.533A.G/p.Asn178Ser (identified in pa- rare variants in CFHR5 associated with IgAN, the functional meaning of these rare tients) in exon 4, c.822A.T/p.Glu274Asp variants was predicted by in silico programs. At last, nine potential functional variants (identified in patients) in exon 6, and fi were identi ed, which were located in different SCRs of the CFHR5 protein. CHB, c.1357C.T/p.Pro453Ser (identified in Chinese Han Beijing; dbSNP, single nucleotide polymorphism database; ExAC, controls) in exon 9. Three variants would exome aggregation consortium; HSF, human splicing finder; SIFT, sorting intolerant lead to protein length change, and they in- from tolerant. cluded c.776T.A/p.Leu259Termination (identified in patients), a stop-gain variant from SKAT, no significant difference was observed regarding that produced a shortened CFHR5 protein (258 amino acids the allele or genotype frequencies of each single common var- only); c.1708T.C/p.Termination570Arg (identified in pa- iant (Table 3). tients and controls), a stop-loss variant that added 37 amino acids to the C terminus of the wild–type CFHR5 protein; and Clinical and Pathologic Manifestations of Patients with c.790+2T.C (identified in patients), a variant located 2 bp IgAN away from an exon-intron boundary (exon 5 to intron 5) that, Among our recruited 500 patients, 43 had rare variants in thus, appeared to break the GT-AG splicing rule; c.790+2T.C CFHR5. Of these patients, only two patients carried two rare was also predicted to be a splicing site change by the Alterna- variants (one patient had c.59–112G.A and c.822A.T, tive Splice Site Predictor (ASSP) and software available from whereas the other patient had c.1147+94G.Tand the Berkeley Drosophila Genome Project (BDGP). Moreover, c.1641G.A), and the remaining 41 patients each carried c.1710*168G.T(identified in patients) in the 39 UTR, located only one rare variant (including c.1147+94G.Tin12pa- 168 bp away from the stop codon, was in the microRNA bind- tients; c.508G.A in seven patients; c.533A.G in six patients; ing region. Therefore, TargetScanHuman 6.2 was used to see

2896 Journal of the American Society of Nephrology J Am Soc Nephrol 27: 2894–2905, 2016 mScNephrol Soc Am J Table 1. Identified genetic variants in CFHR5 MAF dbSNP Position Location Nucleotide Enrolled Individuals 1000 Genomes Phase 3 ExAC Data Version 0.3 (Build142) IgAN, n=500 HC, n=576 Total CHB Total East Asian Chr1:196946445 Promoter c.-350G.T ——0.001 (1/1000) 0.000 (0/1152) 0.000 0.000 0.000 0.000 27: Chr1:196946448 Promoter c.-347G.A — rs9427942 0.000 (0/1000) 0.0009 (1/1152) 0.177 (886/5008) 0.000 (0/206) 0.000 0.000 2894 Chr1:196946869 Intron 1 c.58+17T.A — rs3748557 0.140 (140/1000) 0.141 (163/1152) 0.158 (790/5008) 0.209 (43/206) 0.208 (25,245/121,354) 0.17 (1467/8644) . ——

– Chr1:196946881 Intron 1 c.58+29A C 0.000 (0/1000) 0.0009 (1/1152) 0.000 0.000 0.000 0.000 95 2016 2905, Chr1:196951903 Intron 1 c.59–112G.A ——0.001 (1/1000) 0.000 (0/1152) 0.000 0.000 0.000 0.000 Chr1:196953025 Intron 2 c.254–66A.T — rs570596954 0.002 (2/1000) 0.000 (0/1152) 0.0006 (3/5008) 0.000 (0/206) 0.000 0.000 Chr1:196953169 Exon 3 c.332A.C p.Gln111Pro — 0.001 (1/1000) 0.0009 (1/1152) 0.000 0.000 ,0.0001 (1/121,218) 0.0001 (1/8646) Chr1:196953379 Intron 3 c.430+112T.C ——0.001 (1/1000) 0.000 (0/1152) 0.000 0.000 0.000 0.000 Chr1:196963130 Intron 3 c.431–80G.A — rs117310366 0.013 (13/1000) 0.013 (15/1152) 0.004 (18/5008) 0.029 (6/206) 0.000 0.000 Chr1:196963213 Exon 4 c.434G.A p.Gly145Glu rs57960694 0.013 (13/1000) 0.013 (15/1152) 0.041 (204/5008) 0.029 (6/206) 0.013 (1496/119,312) 0.016 (139/8546) Chr1:196963287 Exon 4 c.508G.A p.Val170Met rs201073457 0.007 (7/1000) 0.004 (5/1152) 0.001 (6/5008) 0.010 (2/206) 0.0006 (77/120,992) 0.008 (70/8644) Chr1:196963312 Exon 4 c.533A.G p.Asn178Ser rs200427185 0.006 (6/1000) 0.000 (0/1152) 0.008 (4/5008) 0.005 (1/206) 0.0003 (37/121,072) 0.004 (37/8646) Chr1:196965015 Exon 5 c.776T.A p.Leu259Termination — 0.001 (1/1000) 0.000 (0/1152) 0.000 0.000 0.000 0.000 Chr1:196965031 Intron 5 c.790+2T.C ——0.001 (1/1000) 0.000 (0/1152) 0.000 0.000 ,0.0001 (1/120,052) 0 (0/8581) Chr1:196965076 Intron 5 c.790+47A.G ——0.001 (1/1000) 0.000 (0/1152) 0.000 0.000 0.000 0.000 Chr1:196965183 Exon 6 c.822A.T p.Glu274Asp — 0.001 (1/1000) 0.000 (0/1152) 0.000 0.000 0.000 0.000 Chr1:196967458 Intron 7 c.1147+24T.G — rs1332664 0.332 (332/1000) 0.316 (364/1152) 0.694 (3475/5008) 0.374 (77/206) 0.752 (87,237/116,050) 0.302 (2560/8467) Chr1:196967528 Intron 7 c.1147+94G.T — rs116891819 0.013 (13/1000) 0.008 (9/1152) 0.004 (18/5008) 0.029 (6/206) 0.000 0.000 Chr1:196971491 Intron 7 c.1148–121G.A ——0.001 (1/1000) 0.000 (0/1152) 0.000 0.000 0.000 0.000 Chr1:196971504 Intron 7 c.1148–108A.G — rs147231103 0.000 (0/1000) 0.009 (10/1152) 0.004 (18/5008) 0.029 (6/206) 0.000 0.000 Chr1:196971544 Intron 7 c.1148–68C.T — rs181464041 0.000 (0/1000) 0.0009 (1/1152) 0.010 (50/5008) 0.000 (0/206) 0.000 0.000 Chr1:196971637 Exon 8 c.1173G.A p.Pro391Pro rs200148491 0.000 (0/1000) 0.0009 (1/1152) 0.0002 (1/5008) 0.000 (0/206) ,0.0001 (5/120,578) 0.0002 (2/8624) Chr1:196971814 Intron 8 c.1330+20G.A ——0.000 (0/1000) 0.0009 (1/1152) 0.000 0.000 ,0.0001 (1/111,288) 0.000 (0/7946) Chr1:196971834 Intron 8 c.1330+40G.A — rs116937944 0.000 (0/1000) 0.0009 (1/1152) 0.0002 (1/5008) 0.005 (1/206) ,0.0001 (2/104,778) 0.000 (0/7380) Chr1:196971872 Intron 8 c.1330+78A.G ——0.001 (1/1000) 0.003 (3/1152) 0.000 0.000 0.000 0.000 Chr1:196973817 Exon 9 c.1357C.T p.Pro453Ser rs184883943 0.000 (0/1000) 0.005 (6/1152) 0.0002 (1/5008) 0.005 (1/206) 0.0001 (13/121,390) 0.002 (13/8654) Chr1:196974101 Intron 9 c.1513+126A.G ——0.003 (3/1000) 0.000 (0/1152) 0.000 0.000 0.000 0.000 Chr1:196974110 Intron 9 c.1513+137G.A — rs142743076 0.000 (0/1000) 0.0009 (1/1152) 0.0002 (1/5008) 0.000 (0/206) 0.000 0.000 Chr1:196977744 Exon 10 c.1641G.A p.Ala547Ala rs74323799 0.001 (1/1000) 0.000 (0/1152) 0.001 (3/5008) 0.000 (0/206) 0.0002 (29/121,322) 0.003 (25/8640) Chr1:196977811 Exon 10 c.1708T.C p.Termination570Arg — 0.001 (1/1000) 0.003 (3/1152) 0.000 0.000 ,0.0001 (5/121,104) 0.0005 (5/8634) (stop codon) Chr1:196977816 39 UTR c.1710*3C.G ——0.001 (1/1000) 0.000 (0/1152) 0.000 0.000 ,0.0001 (9/120,944) 0.000 (0/8626) 9 . —— Chr1:196977981 3 UTR c.1710*168G T 0.001 (1/1000) 0.000 (0/1152) 0.000 0.000 0.000 0.000 www.jasn.org dbSNP, single nucleotide polymorphism database; ExAC, exome aggregation consortium; HC, healthy controls; CHB, Chinese Han Beijing; —, inapplicable. CFHR5 LNCLRESEARCH CLINICAL nIgAN in 2897 CLINICAL RESEARCH www.jasn.org

Table 2. Genetic association analysis to common and rare S: 174.27%69.78%; P,0.001; CFHR5-D: 127.25%61.75%; variants in CFHR5 with IgAN susceptibility using SKAT P,0.001) (Figure 4). However, CFHR5-Short (c.776T.A/p. Test Type N of Variants Q Value P Value Leu259Termination) showed significantly impaired C3b bind- 6 P, Joint 32 4.80 0.03 ing (CFHR5-Short: 56.89% 0.57%; 0.001) (Figure 4). Common only 4 17.31 0.92 Rare only 28 15,240.54 0.002 Q value, test statistic of SKAT. DISCUSSION

On the basis of the GWAS identification of CFH-CFHRsas whether the variant would interfere with microRNA binding factors associated with IgAN predisposition,7,17 the high sim- efficiency. The results showed that c.1710*168G.Tmightalso ilarity of IgAN with CFHR5 nephropathy, and the evidence of be predicted as damaging for its potential influence on the complement activation in IgAN, we suspected that variants in binding of hsa-miR-2117 and hsa-miR-4273 to CFHR5 the CFHR5 gene might also play some part in influencing ge- mRNA. netic predisposition to the complex disease IgAN. In this study, through genetic analysis of the CFHR5 gene, we found Detection of C3b Binding Capacity significant genetic association of rare variants in CFHR5 with Within nine rare variants predicted to influence CFHR5 IgAN, and we further proved an impaired C3b binding capac- function, seven were nonsynonymous. Comparing these seven ity of some rare variants in CFHR5. nonsynonymous variants between patients and controls, we In recent years, increasing evidence indicated that the allelic found four variants that were located in SCR3–5(c.508G.A/p. spectrum of disease-associated variants at a given locus might Val170Met, c.533A.G/p.Asn178Ser, c.776T.A/p.Leu259- include novel, rare, and common variants. Reports from Termination, and c.822A.T/p.Glu274Asp) and showed inflammatory bowel disease and rheumatoid arthritis suppor- higher frequencies in IgAN. Because SCR3–5 in CFHR5 had ted the involvement of both rare and common variants from high amino acid identities similar to SCR10–12 in CFH and single genes or gene sets in common complex diseases.18,19 because SCR12 in CFH has the ability to bind to C3b,16 we Recently, after the identification of membranous nephropathy further evaluated their influence on the recombinant CFHR5 (MN) –associated common variants in PLA2R1 by the GWAS, proteins (Supplemental Figure 1) and its C3b binding capac- Coenen et al.20 convincingly excluded the possibility that rare ity. We found that, when 1–4 mg/ml C3b was added, 1 mg/ml variants in PLA2R1 are involved in MN pathogenesis.21 How- CFHR5 protein (including all of our detected allotypes) was ever, for IgAN, although the GWAS identified several disease– within the linearity range for the C3b binding capacity (Sup- associated loci and common variants,7,17,22,23 few studies have plemental Figure 2). Compared with wild-type CFHR5, focused on rare variants in candidate genes. Here, after screen- CFHR5-M (c.508G.A/p.Val170Met), CFHR5-S (c.533A. ing for variants in the exons, intronic flanking regions, and G/p.Asn178Ser), and CFHR5-D (c.822A.T/p.Glu274Asp) UTRs of CFHR5, we applied a popular rare variant association showed significantly higher capacities for binding to 2 test, SKAT, to evaluate the association between variants in mg/ml C3b (CFHR5-M: 109.67%63.54%; P=0.02; CFHR5- CFHR5 and IgAN.24 SKAT has two distinguishing features

Table 3. Allele and genotype frequencies of common variants of CFHR5 in patients with IgAN and healthy controls Variants Allele Frequencies Genotype Frequencies c.58+17 T.A A T AA + AT TT IgAN 140 (14.00%) 860 (86.00%) 15+110 (25.00%) 375 (75.00%) HC 163 (14.10%) 989 (85.90%) 7+149 (27.10%) 420 (72.90%) P value 0.92 0.44 c.431–80 G.A A G AA + AG GG IgAN 13 (1.30%) 987 (98.70%) 0+13 (2.60%) 487 (97.40%) HC 15 (1.30%) 1137 (98.70%) 0+15 (2.60%) 561 (97.40%) P value .0.99 .0.99 c.434 G.A A G AA + AG GG IgAN 13 (1.30%) 987 (98.70%) 0+13 (2.60%) 487 (97.40%) HC 15 (1.30%) 1137 (98.70%) 0+15 (2.60%) 561 (97.40%) P value .0.99 .0.99 c.1147+24 T.G G T GG + GT TT IgAN 332 (33.20%) 668 (66.80%) 61+210 (54.20%) 229 (45.80%) HC 364 (31.60%) 788 (68.40%) 49+266 (54.70%) 261 (45.30%) P value 0.43 0.87 HC, healthy controls.

2898 Journal of the American Society of Nephrology J Am Soc Nephrol 27: 2894–2905, 2016 www.jasn.org CLINICAL RESEARCH

Table 4. Comparison of clinical and pathologic manifestations in patients with IgAN according to rare variants in CFHR5 Potential Pathogenic Variants Characteristics Total, n=500 P Valuea With, n=43 Without, n=457 Baseline Age, yr, mean6SD 34.23611.09 33.72611.71 34.27611.05 0.75 Men 258 (51.60%) 23 (53.50%) 235 (51.40%) 0.79 Hypertension 0.94 With hypertension 68 (13.62%) 6 (14.00%) 62 (13.60%) Without hypertension 432 (86.38%) 37 (86.00%) 395 (86.40%) Initial proteinuria, g/d, median (IQR) 1.42 (0.81, 2.66) 1.40 (0.83, 2.51) 1.43 (0.80, 2.72) 0.92 ,1 168 (33.60%) 14 (32.60%) 154 (33.70%) 1–3.5 253 (50.60%) 22 (51.20%) 231 (50.50%) $3.5 79 (15.80%) 7 (16.20%) 72 (15.80%) eGFR, ml/min per 1.73 m2, mean6SD 84.19627.03 82.13622.40 84.39627.43 0.60 Haas classification 0.43b 1 43 (8.60%) 3 (7.00%) 40 (8.80%) 2 4 (0.80%) 0 (0.00%) 4 (0.90%) 3 226 (45.20%) 18 (41.90%) 208 (45.50%) 4 166 (33.20%) 16 (37.10%) 150 (32.80%) 5 61 (12.20%) 6 (14.00%) 55 (12.00%) Serum IgA level, g/L, mean6SD 3.3461.22 3.2961.53 3.3561.19 0.77 Serum C3 level, g/L, mean6SD 1.0460.28 1.1160.39 1.0460.27 0.14 IgA deposition (1+/2+/3+–4+) 14 (2.80%)/78 (15.60%)/ 1 (2.30%)/9 (20.90%)/33 13 (2.80%)/69 (15.10%)/ 0.52c 408 (81.60%) (76.70%) 375 (82.10%) C3 deposition (2/1+/2+/3+–4+) 41 (8.20%)/57 (11.40%)/ 5 (11.60%)/5 (11.60%)/ 36 (7.90%)/52 (11.40%)/ 0.11c 164 (32.80%)/238 19 (44.20%)/14 145 (31.70%)/224 (47.60%) (32.60%) (49.00%) Ratio of glomeruli with crescents (IQR) 0.05 (0.00, 0.15) 0.04 (0.00, 0.12) 0.05 (0.00, 0.16) 0.39 Follow-up Follow-up time, mo, mean6SD 56.26636.61 62.51637.35 55.67631.00 0.17 Prednisone/immunosuppressive agents 208 (41.60%) 19 (44.20%) 189 (41.40%) 0.72 ARBs/ACE inhibitors 483 (96.60%) 43 (100%) 440 (96.30%) 0.40 Time average proteinuria, g/d, median (IQR) 0.75 (0.44, 1.41) 0.71 (0.37, 1.70) 0.77 (0.44, 1.38) 0.95 Slope, ml/min per mo, median (IQR) 20.20 (20.51, 0.00) 20.17 (20.53, 20.10) 20.20 (20.51, 0.00) 0.80 Composite events, no. 84 (16.80%) 10 (23.30%) 74 (16.20%) 0.24 IQR, interquartile range; ARB, angiotensin receptor blocker; ACE, angiotensin-converting enzyme. aThe P value was used to indicate the difference between with and without potential pathogenic variants group. bThe P value was calculated between Haas 1–3 and Haas 4–5. cThe P value was calculated by linear-by-linear association. among group–wise association tests. First, SKAT can evaluate to be the only causal factor in the disease. Accordingly, we not only the effect of rare variants but also, the cumulative failed to observe significant differences regarding clinical find- effect of rare and common variants.25 Second, SKATallows for ings, pathologic features, and renal outcomes between patients rare variants to have opposite effects (both effective and del- with or without rare variants in CFHR5. In monogenic dis- eterious) on the traits.26 Because we identified both rare and eases induced by abnormal complement activation, including common variants as well as their gain-of-function or loss-of- aHUS and DDD, in both CFH and CFHR5 were function properties, we tested the genetic associations between reported,27–30 but in complement–involved complex disease, IgAN and rare plus common variants, rare variants alone, and such as SLE, age–related , and IgAN, common variants alone in CFHR5 by SKAT. Different from the although variants in CFH were widely investigated,15,31,32 lim- findings in MN, we found that it was rare variants and not ited studies focused on the CFHR5 gene. To the best of our common variants in CFHR5 that contribute to IgAN suscep- knowledge, our study is the first to report the association of tibility. rare variants in CFHR5 with IgAN. The role of rare variants in In this study, the proportion of patients with IgAN bearing CFHR5 still remains to be elucidated in other complement– rare variants in CFHR5 reached 8.6% (43 of 500). Because involved complex diseases and validated in larger IgAN co- IgAN is widely regarded as a polygenic disease, rare variants horts of other ethnicities. in the CFHR5 gene are likely to play some role in IgAN pre- By in silico prediction, we identified nine potential func- disposition, but they are not, as in CFHR5 nephropathy, likely tional variants of 28 rare variants to influence the CFHR5

J Am Soc Nephrol 27: 2894–2905, 2016 CFHR5 in IgAN 2899 CLINICAL RESEARCH www.jasn.org

+2T.C is a splicing acceptor site variant. The minor allele C would affect the complex dynamics of splicing and can poten- tially lead to a functional change of CFHR5 because of erro- neous exon skipping or the inclusion of a nonintronic sequence. For the other four variants in SCR3–5, we detected their influence on the CFHR5 protein with regard to C3b binding. Results showed that three variant–associated CFHR5 proteins, CFHR5-M (c.508G.A/p.Val170Met), CFHR5-S (c.533A.G/p.Asn178Ser), and CFHR5-D (c.822A.T/p.Glu274Asp), enhanced the binding of CFHR5 to C3b. Recently, CFHR5 was reported to deregulate comple- ment activation by competing with CFH.33,34 Because a high Figure 2. Kaplan–Meier renal survival curves of patients with IgAN according to whether patients were with or without rare degree of complement activation was associated with poor re- variants in CFHR5. In our population of patients with IgAN, event- nal survival in IgAN, it is reasonable for us to suspect CFHR5- free survival of composite outcomes was comparable between M, CFHR5-S, and CFHR5-D as gain of function for CFHR5; patients with rare variants in CFHR5 (n=43) and those without rare therefore, they exhibited stronger competition with CFH, variants (n=457; P value for log-rank test was P=0.92). resulted in a higher degree of complement activation, and contributed to IgAN susceptibility. Different from these, CFHR5-Short resulted from a stop-gain variant (c.776T.A/ effect, although we cannot exclude the possibility of some p.Leu259Termination) and exhibited significantly lower C3b others also having functional meaning. Except for binding capacity. This mutant CFHR5 (CFHR5-Short), as a c.1710*168G.T, which is located in 39 UTR and predicted half-truncated protein (258 amino acids only compared with to change CFHR5 expression, the other eight variants were 569 amino acids in wild-type CFHR5), was prematurely ter- assumed to influence CFHR5 function. SCRs are the basic minated at SCR4, and it, therefore, is missing SCR5–9andpart functional domains in CFH and CFHRs, including of SCR4. Because the accurate mechanism for SCR5–9isstill CFHR5.12 When we compared the allele frequencies of the unknown, we suspected that it might change the biologic eight potential functional variants in patients with IgAN and function of CFHR5 through some other mechanism harbored healthy controls, the interesting thing that we found was that by SCR5–9, which overcame its influence on C3b binding variants that showed higher frequencies in IgAN (c.508G.A, capacity and contributed to IgAN susceptibility. Unfortu- c.533A.G, c.776T.A, c.790+2T.C, and c.822A.T) clus- nately, our study was unable to clarify the pathogenic mechanism tered in SCR3–5, whereas those with higher frequencies in of CFHR5-Short, and therefore, additional studies will be re- controls (c.1357C.T and c.1708T.C) clustered in SCR8–9. quired in the future after the biologic effect of SCR5–9 is revealed. The variant with comparable frequency between patients and The other two potential functional variants, which showed controls (c.332A.C) was located in SCR2. higher frequencies in controls, were located in SCR8–9in In the SCR3–5 region, we identified five potential patho- CFHR5 and correspond to SCR19–20 in CFH.12 SCR19–20 genic variants, with higher frequencies in IgAN. Variant c.790 in CFH is critical for binding to the cell surface. Most genetic mutations in CFH that induced aHUS,35 a disease with severe endothelial injury, are located in SCR19–20.36 Therefore, we as- sumedthatvariantsinCFHR5SCR8–9 would have a similar influence and might be associated with endothelial injury. Al- though we now have considerable evidence regarding endothelial injury in IgAN,37–39 the underlying mechanism is still un- known. Additional functional studies are required to clarify whether our identified variants in SCR8–9ofCFHR5influence its Figure 3. Schematic illustration for the location of nine potential pathogenic variants in binding to endothelial cells and what their CFHR5. The CFHR5 protein contains nine functional domains called SCRs, which are . . genetic mechanisms are in IgAN. drawn as ovals. Variant c.332A C/p.Gln111Pro is located in SCR2, c.508G A/p. fl Val170Met and c.533A.G/p.Asn178Ser are in SCR3, c.776T.A/p.Leu259Termination Different from those variants that in u- . is in SCR4, c.822A.T/p.Glu274Asp in SCR5, and c.1357C.T/p.Pro453Ser and enced CFHR5 function, c.1710*168G T, c.1708T.C/p.Ter570Arg are located in SCR8 and SCR9, respectively. Splicing variant located in the 39 UTR, was predicted to c.790+2T.Cis2bpdownstreamoftheexon5–intron 5 boundary. Variant c.1710* influence the expression rather than the 168G.Tispositionedat39 UTR. sequence of CFHR5. In recent years,

2900 Journal of the American Society of Nephrology J Am Soc Nephrol 27: 2894–2905, 2016 mScNephrol Soc Am J Table 5. Predicted functional rare variants in CFHR5 Identified in Position Location Nucleotide Amino Acid dbSNP (Build142) Function Change Prediction Software Population Chr1:196946445 Promoter c.-350G.T — Patients — Not transcript elements Promoter 2.0 binding sites 27: Chr1:196946448 Promoter c.-347G.A — Controls rs9427942 Not transcript elements Promoter 2.0 2894 binding sites

– . — —

95 2016 2905, Chr1:196946881 Intron 1 c.58+29A C Controls Not branch-point mutation HSF 2.4.1 Chr1:196951903 Intron 1 c.59–112G.A — Patients — Not branch-point mutation HSF 2.4.1 Chr1:196953025 Intron 2 c.254–66A.T — Patients rs570596954 Not branch-point mutation HSF 2.4.1 Chr1:196953169 Exon 3 c.332A.C p.Gln111Pro Patients and controls — Tolerated/benign SIFT-Polyphen-2 Chr1:196953379 Intron 3 c.430+112T.C — Patients — Not branch-point mutation HSF 2.4.1 Chr1:196963287 Exon 4 c.508G.A p.Val170Met Patients and controls rs201073457 Tolerated/probably SIFT-Polyphen-2 damaging Chr1:196963312 Exon 4 c.533A.G p.Asn178Ser Patients rs200427185 Damaging/benign SIFT-Polyphen-2 Chr1:196965015 Exon 5 c.776T.A p.Leu259Termination Patients — Stop gain — Chr1:196965031 Intron 5 c.790+2T.C — Patients — Splicing site BDGP, ASSP Chr1:196965076 Intron 5 c.790+47A.G — Patients — Not branch-point mutation HSF 2.4.1 Chr1:196965183 Exon 6 c.822A.T p.Glu274Asp Patients — Tolerated/benign SIFT-Polyphen-2 Chr1:196967528 Intron 7 c.1147+94G.T — Patients and controls rs116891819 Not branch-point mutation HSF 2.4.1 Chr1:196971491 Intron 7 c.1148–121G.A — Patients — Not branch-point mutation HSF 2.4.1 Chr1:196971504 Intron 7 c.1148–108A.G — Controls rs147231103 Not branch-point mutation HSF 2.4.1 Chr1:196971544 Intron 7 c.1148–68C.T — Controls rs181464041 Not branch-point mutation HSF 2.4.1 Chr1:196971637 Exon 8 c.1173G.A p.Pro391Pro Controls rs200148491 Synonymous mutation — Chr1:196971814 Intron 8 c.1330+20G.A — Controls — Not branch-point mutation HSF 2.4.1 Chr1:196971834 Intron 8 c.1330+40G.A — Controls rs116937944 Not branch-point mutation HSF 2.4.1 Chr1:196971872 Intron 8 c.1330+78A.G — Patients and controls — Not branch-point mutation HSF 2.4.1 Chr1:196973817 Exon 9 c.1357C.T p.Pro453Ser Controls rs184883943 Tolerated/probably SIFT-Polyphen-2 damaging Chr1:196974101 Intron 9 c.1513+126A.G — Patients — Not branch-point mutation HSF 2.4.1 Chr1:196974110 Intron 9 c.1513+137G.A — Controls rs142743076 Not branch-point mutation HSF Chr1:196977744 Exon 10 c.1641G.A p.Ala547Ala Patients rs74323799 Synonymous mutation —

Chr1:196977811 Exon 10 (stop codon) c.1708T.C p.Termination570Arg Patients and controls — Stop loss — www.jasn.org Chr1:196977816 39 UTR c.1710*3C.G — Patients — Not microRNA binding site TargetScanHuman 6.2 Chr1:196977981 39 UTR c.1710*168G.T — Patients — hsa-mir-2117/hsa-mir-4273 TargetScanHuman 6.2 binding site dbSNP, single nucleotide polymorphism database; —, inapplicable; HSF, Human Splicing Finder; SIFT, Sorting Intolerant from Tolerant; Polyphen-2, Polymorphism Phenotyping v2. CFHR5 LNCLRESEARCH CLINICAL nIgAN in 2901 CLINICAL RESEARCH www.jasn.org

to IgAN predisposition through its post-transcriptional regula- tion of CFHR5 expression. Our study had several limitations. First, we identified several potential functional rare variants. However, C3b binding capacity was only detected in some of them. Additional functional studies are required to reveal their exact biologic roles in IgAN. Second, because we failed to harvest the secreted CFHR5 from an insect cell system, an Escherichia coli system, which lacks post-translational modifications, was used for recombinant CFHR5 protein expression. Whether post- translational modification would influence the C3b binding capacity of CFHR5 is still unknown. Third, diverse incidences Figure 4. Rare coding variants induced a change in the C3b of IgAN were reported in different racial and ethnic popula- binding capacity of the CFHR5 protein. Compared with CFHR5- tions, and therefore, future replication studies are important CFHR5 WT, CFHR5-M (c.508G.A/p.Val170Met), CFHR5-S (c.533A.G/p. to validate the genetic association of in IgAN. We failed Asn178Ser), and CFHR5-D (c.822A.T/p.Glu274Asp) showed to find any genotype-phenotype correlation for rare variants significantly higher binding capacities for C3b (CFHR5-M: in CFHR5. Whether rare variants in CFHR5 contributed to 109.67%63.54%; P=0.02; CFHR5-S: 174.27%69.78%; P,0.001; clinical or pathologic phenotypes or the progression of IgAN CFHR5-D: 127.25%61.75%; P,0.001), whereas CFHR5-Short is a question awaiting further convincing evidence. (c.776T.A/p.Leu259Termination) had a significantly lower C3b In conclusion, through screening the genetic variants in 6 , binding capacity (CFHR5-Short: 56.89% 0.57%; P 0.001). Ex- CFHR5, we found that rare variants in CFHR5 contributed to periments were performed in triplicate and repeated three times genetic susceptibility to IgAN, which suggested that CFHR5 is with similar results. White columns represent quantities of re- an IgAN susceptibility gene. combinant CFHR5 proteins coated on the plates. Black columns represent C3b binding capacity of the coated CFHR5 proteins compared with wild-type CFHR5. Blank-1, no CFHR5 + C3b + polyclonal anti–C3c/CFHR5 ; blank-2, no CFHR5 + no CONCISE METHODS C3b + polyclonal anti–C3c/CFHR5 antibody. Study Population Here, we recruited 1076 northern Chinese individuals, including 500 increasing microRNA–associated single–nucleotide polymor- patients with sporadic IgAN (diagnosed from 2003 to 2009) and 576 phisms were found on microRNA target sites within the 39 age-, sex-, and geographically matched unrelated healthy controls UTRs of mRNAs, such as hsa-miR-1207–5p in CFHR5 nephrop- during this time interval. The diagnosis of IgAN was proven by renal athy,40 miR-148b in IgAN,41 and so on. The c.1710*168G.T biopsy on the basis of granular deposition of IgA in the glomerular variant in CFHR5 was at the target site for hsa-miR-2117/has- mesangium by immunofluorescence detection and the deposition of miR-4273 binding. The mutant Tallele substituted for the G allele electron-dense material in the mesangium by ultrastructural exam- theoretically influences the binding of miR-2117/miR-4273 and ination. Patients with Henoch–Schonlein purpura, SLE, and chronic therefore, would change CFHR5 levels. It was reported that hepatic diseases were excluded by detailed clinical and laboratory CFHR5 colocalized with complement–containing glomerular examinations. immune deposits in a variety of glomerular pathologic states.42 The study protocol was approved by the Medical Ethics Committee Although it has an unclear biologic role, CFHR5 is postulated to of Peking University, and informed written consent was obtained from have a physiologic role in complement processing within the all patients. Figure 1 shows the flow chart of genetic analysis applied in kidney. We speculated that c.1710*168G.T might contribute our study.

Table 6. PCR conditions and primers used to amplify CFHR5 gene Exon Forward (59–39) Reverse (59–39) Annealing Temperature, °C Fragment, bp 1 CTTGCTTGCCTTTTGAAACA CCCCTTCAAATTATCCTCAGC 61.2 585 2 CTGGGCAACAAGAGTAAAACCT TCTCAAAATAGGAGGACTACATCTC 61.2 582 3 CGGTAGCATGACCCAAATTC GGTAGGCAAACTATGTTATTGCAC 61.2 488 4 AGTTTCCCAATTTGCCTGAG CTGCATCCTTTCTCCTTTGC 61.2 501 5 and 6 GAGGAAACGAATGCAGTCAA TCCATTCATCATGCCAGAAA 57 901 7 TCAGTCAAAACTCCCACTAGGAA CATCTTTACCAGAAAGCCAAGG 57 637 8 GGAGATACAAGAGAGCATCTGAAA GTTTCTTCTATGAACACTGTTGGAG 61.2 636 9 AATTATTTGAATTTCCAGACACCTT CGAATAGGCCCCATAAATAGC 57 568 10 CATATGTAGCCCATACACAGTGC CCCCACCATCTTGGACTTCT 61.2 690

2902 Journal of the American Society of Nephrology J Am Soc Nephrol 27: 2894–2905, 2016 www.jasn.org CLINICAL RESEARCH

Screening for Genetic Variants in CFHR5 Computational Assessment of Rare Variants in CFHR5 Genomic DNAwas extracted from EDTAanticoagulated whole–blood A functional evaluation of the identified rare variants in CFHR5 was samples by a salting out procedure.43 The genomic DNA reference analyzed using several in silico prediction programs according to their sequence (version NC_000001.11) and mRNA reference sequence position in the genome. For variants in the promoter and 59 UTR, (version NM_030787.3) for the CFHR5 gene were obtained from promoter 2.0 (http://www.cbs.dtu.dk/services/Promoter/) was used the gene database (http://www.ncbi.nlm.nih.gov/entrez/ to predict whether the variants were in transcription element binding query.fcgi?db=gene). For the enrolled 500 patients with IgAN and sites; for those in the coding region, their potential pathogenicity was 576 healthy controls, a coding sequence in each exon with the intronic analyzed by two independent programs: one was Sorting Intolerant flanking regions as well as the 59 and 39 UTRs of the CFHR5 gene were from Tolerant (http://sift.jcvi.org/), with prediction that is on the amplified by PCR. Primers used for PCR were designed by the Primer basis of the degree of the conservation of amino acid residues in 3 program (http://bioinfo.ut.ee/primer3-0.4.0/) and are listed in Ta- sequence alignments derived from closely related sequences, and ble 6 along with the PCR annealing temperatures. PCR products were the other was Polymorphism Phenotyping v2 (http://genetics.bwh. sequenced by the ABI 3730 XL Genetic Analyzer (Applied Biosystems, harvard.edu/pph2/), in which variants with scores of .2.0, 1.5–2.0, Foster City, CA) and aligned to the reference sequence to identify and ,1.5 were defined as probably damaging, possibly damaging, genetic variants in CFHR5.Theidentified genetic variants were and benign, respectively. For variants in intronic flanking region, checked against the dbSNP database (build142) to see if they were splice donor sites or acceptor sites were checked by the splice site reported before. Meanwhile, these variants were also screened against prediction software available by the BDGP and the ASSP, which is a two publicly available databases of exome and genome sequences: the sequence analysis tool for the prediction and classification of splice 1000 Genomes Project (phase 3; http://www.1000genomes.org) and sites. Meanwhile, splicing branch point A was predicted by Human ExAC browser (Beta; version 0.3; http://exac.broadinstitute.org/). Splicing Finder software (version 2.4.1; http://www.umd.be/HSF/), which helps with the study of the pre-mRNA splicing; for variants in 9 Association Analyses of Variants in CFHR5 with IgAN 3 UTR, a microRNA target prediction tool TargetScanHuman 6.2 To evaluate the genetic contribution of variants in CFHR5 to IgAN (http://www.targetscan.org/), which predicts biologic targets of mi- susceptibility, a popular genetic association test, SKAT, was applied. croRNAs by searching for the presence of conserved 8mer and 7mer The R package SKAT 1.0.7 was used for the analysis. At first, the sites that match the seed region of each microRNA, was used to combined effect of all variants in CFHR5 was analyzed. Then, variants evaluate their influence on microRNA binding. in CFHR5 were grossly divided into common and rare variants ac- cording to their MAF in healthy controls (cutoff =1%). The effects of Detection of the Influence of Rare Variants on Binding rare variants alone and common variants alone were evaluated by Capacity of CFHR5 to C3b SKAT_CommonRare. Additionally, for common variants in Recombinant CFHR5 proteins (including the wild type [CFHR5-WT] CFHR5, comparisons of allele and genotype frequencies at the sin- and those containing p.Val170Met [CFHR5-M], p.Asn178Ser gle-variant level were also performed between 500 patients with IgAN [CFHR5-S], p.Glu274Asp [CFHR5-D], and p.Leu259Termination and 576 healthy controls. [CFHR5-Short]) were supplied by GenScript Corporation through customization. In brief, codon–optimized DNA sequences coding for Comparative Analyses between Patients with IgAN CFHR5 as well as eight His amino acids at the C terminus of CFHR5, According to Rare Variants in CFHR5 were synthesized and ligated into a pUC57 vector and expressed in E. For the enrolled 500 patients with sporadic IgAN, clinical parameters coli expression systems. Recombinant CFHR5 proteins were purified at the time of renal biopsy, including serum IgA level, serum C3 level, by His tag affinity chromatography from the culture medium of E. serum creatinine level, 24-hour urine protein excretion, and history of coli. For the detection of the binding capacity of recombinant CFHR5 hypertension, were collected from the medical records. The eGFR was proteins to C3b, recombinant CFHR5 (1 mg/ml) was coated onto calculated using the Modified Glomerular Filtration Rate Estimating MaxiSorp Plastic Plates (Nalge-Nunc, Rochester, NY) at 4°C over- Equation for Chinese.44 The Haas classification was assessed in pa- night. After being blocked with 1% BSA (Sigma-Aldrich, St. Louis, tients whose renal biopsy tissues contained more than eight glomer- MO), 2 mg/ml C3b (Calbiochem, San Diego, CA) was added, and the uli.45 mixture was incubated for 1 hour at 37°C. The binding of C3b to According to the presence or absence of rare variants in CFHR5, CFHR5 was examined by Polyclonal Rabbit Anti-Human C3c Anti- patients with IgAN were divided into two groups: those with rare body (DAKO, Copenhagen, Denmark) and followed by horseradish variants and those without rare variants. At first, baseline clinical peroxidase–labeled donkey anti–rabbit IgG (Santa Cruz Biotechnol- manifestations and histologic classifications were compared between ogy, Santa Cruz, CA). The reaction was developed using 3,39,5,59- the two groups. Then, survival analysis was used to evaluate whether Tetramethylbenzidine Horseradish Peroxidase Color Development patients in the two groups had a significant difference in long–term Solution for ELISA (Sigma-Aldrich) and stopped with 1 mol/L sul- renal outcomes. A composite end point, which was defined as 50% furic acid before the absorbance was measured at 450/570 nm. eGFR decline, ESRD, or death, whichever occurred first, was used in Meanwhile, 0.125 mg/ml recombinant CFHR5 was also coated and this study. ESRD was defined as eGFR,15 ml/min per 1.73 m2 or the blocked as described above, and polyclonal anti–CFHR5 antibody need for RRT (such as hemodialysis, peritoneal dialysis, or renal (ab175254; Abcam, Inc., Cambridge, MA) and horseradish peroxi- transplantation) for the purpose of this study. dase–labeled goat anti–rabbit IgG (Santa Cruz Biotechnology) were

J Am Soc Nephrol 27: 2894–2905, 2016 CFHR5 in IgAN 2903 CLINICAL RESEARCH www.jasn.org sequentially added and incubated for 1 hour at 37°C. After reaction R, Izzi C, Viola BF, Prati E, Salvadori M, Mignani R, Gesualdo L, development and stopping with 1 mol/L sulfuric acid, the absorbance Bertinetto F, Mesiano P, Amoroso A, Scolari F, Chen N, Zhang H, Lifton fi was measured at 450/570 nm. RP: Genome-wide association study identi es susceptibility loci for IgA nephropathy. Nat Genet 43: 321–327, 2011 8. Kiryluk K, Li Y, Sanna-Cherchi S, Rohanizadegan M, Suzuki H, Eitner F, Statistical Analyses Snyder HJ, Choi M, Hou P, Scolari F, Izzi C, Gigante M, Gesualdo L, Statistical analyses were performed using SPSS software (version 16.0; Savoldi S, Amoroso A, Cusi D, Zamboli P, Julian BA, Novak J, Wyatt RJ, SPSS Inc., Chicago, IL). Normally distributed quantitative variables Mucha K, Perola M, Kristiansson K, Viktorin A, Magnusson PK, were expressed as the means6SDs, whereas for those with a non- Thorleifsson G, Thorsteinsdottir U, Stefansson K, Boland A, Metzger M, ThibaudinL,WannerC,JagerKJ,GotoS,MaixnerovaD,KarnibHH, normal distribution, the medians and interquartile ranges were used. Nagy J, Panzer U, Xie J, Chen N, Tesar V, Narita I, Berthoux F, Floege J, Categorical data were summarized as absolute frequencies and per- Stengel B, Zhang H, Lifton RP, Gharavi AG: Geographic differences in centages. For continuous variables, the independent samples t test genetic susceptibility to IgA nephropathy: GWAS replication study and was used if the data had a normal distribution, and if not, the Mann– geospatial risk analysis. PLoS Genet 8: e1002765, 2012 Whitney or Kruskal–Wallis test was performed. Categorical variables 9. Skerka C, Chen Q, Fremeaux-Bacchi V, Roumenina LT: Complement factor H related proteins (CFHRs). Mol Immunol 56: 170–180, 2013 were compared using the chi-squared test. Cumulative kidney sur- 10. Gale DP, de Jorge EG, Cook HT, Martinez-Barricarte R, Hadjisavvas A, vival curves were derived using the Kaplan–Meier method, and dif- McLean AG, Pusey CD, Pierides A, Kyriacou K, Athanasiou Y, ferences between the curves were analyzed using the log-rank test. A Voskarides K, Deltas C, Palmer A, Frémeaux-Bacchi V, de Cordoba SR, two-tailed P,0.05 was considered statistically significant. Maxwell PH, Pickering MC: Identification of a mutation in complement factor H-related protein 5 in patients of Cypriot origin with glomeru- lonephritis. Lancet 376: 794–801, 2010 11. Barbour TD, Pickering MC, Cook HT: Recent insights into C3 glomer- ulopathy. Nephrol Dial Transplant 28: 1685–1693, 2013 ACKNOWLEDGMENTS 12.McRaeJL,CowanPJ,PowerDA,MitchelhillKI,KempBE,Morgan BP, Murphy BF: Human factor H-related protein 5 (FHR-5). A new This work was supported by Major State Basic Research Development complement-associated protein. JBiolChem276: 6747–6754, Program of China 973 Program grant 2012CB517700, Natural Science 2001 13. McRae JL, Duthy TG, Griggs KM, Ormsby RJ, Cowan PJ, Cromer BA, Foundation for Innovation Research Group of China grant 81321064, McKinstry WJ, Parker MW, Murphy BF, Gordon DL: Human factor H- Capital of Clinical Characteristics and Applied Research Fund grant related protein 5 has cofactor activity, inhibits C3 convertase activity, Z141107002514037, National Science Foundation of China grant binds heparin and C-reactive protein, and associates with lipoprotein. J 81470945, and Beijing Natural Science Foundation grant 7131016. Immunol 174: 6250–6256, 2005 14. Kidney Disease: Improving Global Outcomes (KDIGO) Glomerulone- phritis Work Group: KDIGO Clinical Practice Guideline for Glomeru- lonephritis. Kidney Int Suppl 2: 209–217, 2012 DISCLOSURES 15. Zhu L, Zhai YL, Wang FM, Hou P, Lv JC, Xu DM, Shi SF, Liu LJ, Yu F, Zhao None. MH, Novak J, Gharavi AG, Zhang H: Variants in complement factor H and complement factor H-related protein genes, CFHR3 and CFHR1, affect complement activation in IgA nephropathy. J Am Soc Nephrol 26: 1195–1204, 2015 REFERENCES 16. Jokiranta TS, Hellwage J, Koistinen V, Zipfel PF, Meri S: Each of the three binding sites on complement factor H interacts with a distinct site 1. D’Amico G: The commonest glomerulonephritis in the world: IgA ne- on C3b. JBiolChem275: 27657–27662, 2000 phropathy. QJMed64: 709–727, 1987 17. Yu XQ, Li M, Zhang H, Low HQ, Wei X, Wang JQ, Sun LD, Sim KS, Li Y, 2. Allen AC, Bailey EM, Brenchley PE, Buck KS, Barratt J, Feehally J: Foo JN, Wang W, Li ZJ, Yin XY, Tang XQ, Fan L, Chen J, Li RS, Wan JX, Mesangial IgA1 in IgA nephropathy exhibits aberrant O-glycosylation: Liu ZS, Lou TQ, Zhu L, Huang XJ, Zhang XJ, Liu ZH, Liu JJ: A genome- Observations in three patients. Kidney Int 60: 969–973, 2001 wide association study in Han Chinese identifies multiple susceptibility 3. Suzuki H, Moldoveanu Z, Hall S, Brown R, Julian BA, Wyatt RJ, Tomana loci for IgA nephropathy. Nat Genet 44: 178–182, 2012 M, Tomino Y, Novak J, Mestecky J: IgA nephropathy: Characterization 18. Diogo D, Kurreeman F, Stahl EA, Liao KP, Gupta N, Greenberg JD, of IgG antibodies specificforgalactose-deficient IgA1. Contrib Rivas MA, Hickey B, Flannick J, Thomson B, Guiducci C, Ripke S, Nephrol 157: 129–133, 2007 Adzhubey I, Barton A, Kremer JM, Alfredsson L, Sunyaev S, Martin J, 4. Moura IC, Centelles MN, Arcos-Fajardo M, Malheiros DM, Collawn JF, Zhernakova A, Bowes J, Eyre S, Siminovitch KA, Gregersen PK, Cooper MD, Monteiro RC: Identification of the transferrin receptor as a Worthington J, Klareskog L, Padyukov L, Raychaudhuri S, Plenge RM; novel immunoglobulin (Ig)A1 receptor and its enhanced expression on Consortium of Rheumatology Researchers of North America; Rheu- mesangial cells in IgA nephropathy. JExpMed194: 417–425, 2001 matoid Arthritis Consortium International: Rare, low-frequency, and 5. Lai KN, Tang SC, Leung JC: Recent advances in IgA nephropathy–the common variants in the protein-coding sequence of biological candi- glomerulopodocytic-tubular communication. Adv Otorhinolaryngol date genes from GWASs contribute to risk of rheumatoid arthritis. Am J 72: 40–44, 2011 Hum Genet 92: 15–27, 2013 6. Haas M: Histology and immunohistology of IgA nephropathy. J 19. Rivas MA, Beaudoin M, Gardet A, Stevens C, Sharma Y, Zhang CK, Nephrol 18: 676–680, 2005 Boucher G, Ripke S, Ellinghaus D, Burtt N, Fennell T, Kirby A, Latiano A, 7. Gharavi AG, Kiryluk K, Choi M, Li Y, Hou P, Xie J, Sanna-Cherchi S, Men Goyette P, Green T, Halfvarson J, Haritunians T, Korn JM, Kuruvilla F, CJ, Julian BA, Wyatt RJ, Novak J, He JC, Wang H, Lv J, Zhu L, Wang W, Lagacé C, Neale B, Lo KS, Schumm P, Törkvist L, Dubinsky MC, Brant SR, Wang Z, Yasuno K, Gunel M, Mane S, Umlauf S, Tikhonova I, Beerman I, Silverberg MS, Duerr RH, Altshuler D, Gabriel S, Lettre G, Franke A, Savoldi S, Magistroni R, Ghiggeri GM, Bodria M, Lugani F, Ravani P, D’Amato M, McGovern DP, Cho JH, Rioux JD, Xavier RJ, Daly MJ; Na- Ponticelli C, Allegri L, Boscutti G, Frasca G, Amore A, Peruzzi L, Coppo tional Institute of Diabetes and Digestive Kidney Diseases Inflammatory

2904 Journal of the American Society of Nephrology J Am Soc Nephrol 27: 2894–2905, 2016 www.jasn.org CLINICAL RESEARCH

Bowel Disease Genetics Consortium (NIDDK IBDGC); United Kingdom 31. Jönsen A, Nilsson SC, Ahlqvist E, Svenungsson E, Gunnarsson I, Inflammatory Bowel Disease Genetics Consortium; International In- Eriksson KG, Bengtsson A, Zickert A, Eloranta ML, Truedsson L, flammatory Bowel Disease Genetics Consortium: Deep resequencing Rönnblom L, Nordmark G, Sturfelt G, Blom AM: Mutations in genes of GWAS loci identifies independent rare variants associated with in- encoding complement inhibitors CD46 and CFH affect the age at ne- flammatory bowel disease. Nat Genet 43: 1066–1073, 2011 phritis onset in patients with systemic lupus erythematosus. Arthritis 20. Coenen MJ, Hofstra JM, Debiec H, Stanescu HC, Medlar AJ, Stengel B, Res Ther 13: R206, 2011 Boland-Augé A, Groothuismink JM, Bockenhauer D, Powis SH, 32. Narendra U, Pauer GJ, Hagstrom SA: Genetic analysis of complement Mathieson PW, Brenchley PE, Kleta R, Wetzels JF, Ronco P: Phospho- factor H related 5, CFHR5, in patients with age-related macular de- lipase A2 receptor (PLA2R1) sequence variants in idiopathic membra- generation. Mol Vis 15: 731–736, 2009 nous nephropathy. JAmSocNephrol24: 677–683, 2013 33. Goicoechea de Jorge E, Caesar JJ, Malik TH, Patel M, Colledge M, 21. Salant DJ: Genetic variants in membranous nephropathy: Perhaps a Johnson S, Hakobyan S, Morgan BP, Harris CL, Pickering MC, Lea SM: perfect storm rather than a straightforward conformeropathy? JAmSoc Dimerization of complement factor H-related proteins modulates com- Nephrol 24: 525–528, 2013 plement activation in vivo. Proc Natl Acad Sci U S A 110: 4685–4690, 2013 22. Feehally J, Farrall M, Boland A, Gale DP, Gut I, Heath S, Kumar A, Peden JF, 34. Tortajada A, Yébenes H, Abarrategui-Garrido C, Anter J, García- Maxwell PH, Morris DL, Padmanabhan S, Vyse TJ, Zawadzka A, Rees AJ, Fernández JM, Martínez-Barricarte R, Alba-Domínguez M, Malik TH, Lathrop M, Ratcliffe PJ: HLA has strongest association with IgA nephrop- Bedoya R, Cabrera Pérez R, López Trascasa M, Pickering MC, Harris CL, athy in genome-wide analysis. J Am Soc Nephrol 21: 1791–1797, 2010 Sánchez-Corral P, Llorca O, Rodríguez de Córdoba S: C3 glomerul- 23. Kiryluk K, Li Y, Scolari F, Sanna-Cherchi S, Choi M, Verbitsky M, Fasel D, opathy-associated CFHR1 mutation alters FHR oligomerization and Lata S, Prakash S, Shapiro S, Fischman C, Snyder HJ, Appel G, Izzi C, complement regulation. JClinInvest123: 2434–2446, 2013 Viola BF, Dallera N, Del Vecchio L, Barlassina C, Salvi E, Bertinetto FE, 35. Kavanagh D, Goodship T: Genetics and complement in atypical HUS. Amoroso A, Savoldi S, Rocchietti M, Amore A, Peruzzi L, Coppo R, Pediatr Nephrol 25: 2431–2442, 2010 Salvadori M, Ravani P, Magistroni R, Ghiggeri GM, Caridi G, Bodria M, 36. Mele C, Remuzzi G, Noris M: Hemolytic uremic syndrome. Semin Im- Lugani F, Allegri L, Delsante M, Maiorana M, Magnano A, Frasca G, munopathol 36: 399–420, 2014 Boer E, Boscutti G, Ponticelli C, Mignani R, Marcantoni C, Di Landro D, 37. Zhu L, Shi S, Liu L, Lv J, Zhang H: Increased plasma sVCAM-1 is asso- Santoro D, Pani A, Polci R, Feriozzi S, Chicca S, Galliani M, Gigante M, ciated with severity in IgA nephropathy. BMC Nephrol 14: 21, 2013 Gesualdo L, Zamboli P, Battaglia GG, Garozzo M, Maixnerová D, Tesar 38. Zhai YL, Zhu L, Shi SF, Liu LJ, Lv JC, Zhang H: Elevated soluble VEGF V, Eitner F, Rauen T, Floege J, Kovacs T, Nagy J, Mucha K, Pączek L, receptor sFlt-1 correlates with endothelial injury in IgA nephropathy. Zaniew M, Mizerska-Wasiak M, Roszkowska-Blaim M, Pawlaczyk K, Gale PLoS One 9: e101779, 2014 D, Barratt J, Thibaudin L, Berthoux F, Canaud G, Boland A, Metzger M, 39. Zhang JJ, Jiang L, Liu G, Wang SX, Zou WZ, Zhang H, Zhao MH: Ele- Panzer U, Suzuki H, Goto S, Narita I, Caliskan Y, Xie J, Hou P, Chen N, vation of serum von Willebrand factor and anti-endothelial cell anti- Zhang H, Wyatt RJ, Novak J, Julian BA, Feehally J, Stengel B, Cusi D, bodies in patients with immunoglobulin A nephropathy are associated Lifton RP, Gharavi AG: Discovery of new risk loci for IgA nephropathy with intrarenal arterial lesions. Nephrology (Carlton) 13: 712–720, 2008 implicates genes involved in immunity against intestinal . 40. Papagregoriou G, Erguler K, Dweep H, Voskarides K, Koupepidou P, Nat Genet 46: 1187–1196, 2014 Athanasiou Y, Pierides A, Gretz N, Felekkis KN, Deltas C: A miR-1207- 24. Wu MC, Lee S, Cai T, Li Y, Boehnke M, Lin X: Rare-variant association 5p binding site polymorphism abolishes regulation of HBEGF and is testing for sequencing data with the sequence kernel association test. associated with disease severity in CFHR5 nephropathy. PLoS One 7: Am J Hum Genet 89: 82–93, 2011 e31021, 2012 25. Ionita-Laza I, Lee S, Makarov V, Buxbaum JD, Lin X: Sequence kernel 41. Serino G, Sallustio F, Cox SN, Pesce F, Schena FP: Abnormal miR-148b association tests for the combined effect of rare and common variants. expression promotes aberrant glycosylation of IgA1 in IgA nephropa- Am J Hum Genet 92: 841–853, 2013 thy. JAmSocNephrol23: 814–824, 2012 26. Lee S, Wu MC, Lin X: Optimal tests for rare variant effects in sequencing 42. Murphy B, Georgiou T, Machet D, Hill P, McRae J: Factor H-related association studies. Biostatistics 13: 762–775, 2012 protein-5: A novel component of human glomerular immune deposits. 27. Monteferrante G, Brioschi S, Caprioli J, Pianetti G, Bettinaglio P, Bresin Am J Kidney Dis 39: 24–27, 2002 E, Remuzzi G, Noris M: Genetic analysis of the complement factor H 43. Miller SA, Dykes DD, Polesky HF: A simple salting out procedure for related 5 gene in haemolytic uraemic syndrome. Mol Immunol 44: extracting DNA from human nucleated cells. Nucleic Acids Res 16: 1704–1708, 2007 1215, 1988 28. Abrera-Abeleda MA, Nishimura C, Smith JL, Sethi S, McRae JL, Murphy 44. Ma YC, Zuo L, Chen JH, Luo Q, Yu XQ, Li Y, Xu JS, Huang SM, Wang LN, BF, Silvestri G, Skerka C, Józsi M, Zipfel PF, Hageman GS, Smith RJ: Huang W, Wang M, Xu GB, Wang HY: Modified glomerular filtration Variations in the complement regulatory genes factor H (CFH) and rate estimating equation for Chinese patients with chronic kidney dis- factor H related 5 (CFHR5) are associated with membranoproliferative ease. JAmSocNephrol17: 2937–2944, 2006 glomerulonephritis type II (dense deposit disease). J Med Genet 43: 45. Haas M: Histologic subclassification of IgA nephropathy: A clinico- 582–589, 2006 pathologic study of 244 cases. AmJKidneyDis29: 829–842, 1997 29. Chen Q, Wiesener M, Eberhardt HU, Hartmann A, Uzonyi B, Kirschfink M,AmannK,BuettnerM,GoodshipT,HugoC,SkerkaC,ZipfelPF: Complement factor H-related hybrid protein deregulates complement in dense deposit disease. JClinInvest124: 145–155, 2014 See related editorial, “Challenges in Rare Variant Association Studies for 30. Westra D, Vernon KA, Volokhina EB, Pickering MC, van de Kar NC, van Complex Kidney Traits: CFHR5 and IgA Nephropathy,” on pages 2547–2551. den Heuvel LP: Atypical hemolytic uremic syndrome and genetic ab- errations in the complement factor H-related 5 gene. J Hum Genet 57: This article contains supplemental material online at http://jasn.asnjournals. 459–464, 2012 org/lookup/suppl/doi:10.1681/ASN.2015010012/-/DCSupplemental.

J Am Soc Nephrol 27: 2894–2905, 2016 CFHR5 in IgAN 2905 Supplemental Figure Legends

Supplemental Figure 1. Quality control of recombinant CFHR5 proteins by

Western Blot. Recombinant CFHR5 proteins were separated by SDS-PAGE and

transferred to a membrane, the recombinant protein can be detected by anti-6X His

tag antibody (a) and polyclonal anti-CFHR5 antibody (b) to form a band at about 63

kDa (CFHR5-WT, CFHR5-M, CFHR5-D, CFHR5-S) and 28 kDa (CFHR5-Short).

Supplemental Figure 2. C3b binding capacity assay of different allotypes of

CFHR5 proteins with respect to different C3b concentration. The linearity range

for C3b binding capacity of 1μg/mL CFHR5 was 1-4 μg/mL. Among this range,

CFHR5-M (c.508G>A/p.Val170Met), CFHR5-S (c.533A>G/p.Asn178Ser) and

CFHR5-D (c.822A>T/p.Glu274Asp) showed consistent higher binding capacity to

C3b, while CFHR5-Short (c.776T>A/p.Leu259Ter) had lower C3b binding capacity, compared to CFHR5-WT protein. Supplemental Figure 1.

Supplemental Figure 2.