Atypical Hemolytic Uremic Syndrome Formation of the C3 Convertase ...

The Journal of Immunology

A Novel C3 Mutation Causing Increased Formation of the C3 Convertase in Familial Atypical Hemolytic Uremic Syndrome Lisa Sartz,*,1 Anders I. Olin,†,1 Ann-Charlotte Kristoffersson,* Anne-lie Sta˚hl,* Martin E. Johansson,‡,x Kerstin Westman,{ Veronique Fremeaux-Bacchi,‖ Kristina Nilsson-Ekdahl,#,** and Diana Karpman* Atypical hemolytic uremic syndrome has been associated with dysregulation of the alternative complement pathway. In this study, a novel heterozygous C3 mutation was identified in a factor B-binding region in exon 41, V1636A (4973 T > C). The mutation was found in three family members affected with late-onset atypical hemolytic uremic syndrome and symptoms of glomerulonephritis. All three patients exhibited increased complement activation detected by decreased C3 levels and glomerular C3 deposits. Platelets from two of the patients had C3 and C9 deposits on the cell surface. Patient sera exhibited more C3 cleavage and higher levels of C3a. The C3 mutation resulted in increased C3 binding to factor B and increased net formation of the C3 convertase, even after decay induced by decay-accelerating factor and factor H, as assayed by surface plasmon resonance. Patient sera incubated with washed human platelets induced more C3 and C9 deposition on the cell surface in comparison with normal sera. More C3a was released into serum over time when washed platelets were exposed to patient sera. Results regarding C3 and C9 deposition on washed platelets were confirmed using purified patient C3 in C3-depleted serum. The results indicated enhanced convertase formation leading to increased complement activation on cell surfaces. Previously described C3 mutations showed loss of function with regard to C3 binding to complement regulators. To our knowledge, this study presents the first known C3 mutation inducing increased formation of the C3 convertase, thus explaining enhanced activation of the alternative pathway of complement. The Journal of Immunology, 2012, 188: 2030–2037. emolytic uremic syndrome (HUS) is defined as a triad of nonimmune microangiopathic hemolytic anemia, thrombocytopenia, and renal failure. A subtype of HUS, atypical HUS (aHUS) is, in many cases, associated with activation of the alternative pathway of complement. Activation has been linked to mutations in complement factors in 50–60% of cases (1). Mutations have been identified in regulators of the alternative pathway, such as factor H (CFH) (2), factor I (CFI) (3), and membrane-cofactor protein (MCP/CD46) (4), as well as deletions of factor H-related proteins 1 and 3 (CFHR1/3), with the latter being associated with anti-factor H Abs in some patients (5). In addition, mutations have been demonstrated in complement factors C3 (6, 7) and factor B (CFB) (8), as well as in thrombomodulin (9). Up to 10% of aHUS patients have mutations affecting more than one protein (10, 11).

H

Studies showed that many of the identified aHUS-associated mutations affect protein function, causing loss-of-function in complement regulators or gain-of-function in CFB, thus enabling uninhibited complement activation to occur on cell surfaces (1). C3 mutations were described in a cohort of patients with aHUS, and functional studies showed that the mutations resulted in decreased secretion of mutant constructs or in a loss-of-function, demonstrated as decreased C3 binding to MCP or CFH (6). In addition, incubation of C3 mutants with CFI and its cofactor MCP or, to a lesser degree, CFH, exhibited decreased cofactor activity, which would lead to less C3b inactivation. These interactions could explain increased complement activation via the alternative pathway. However, none of the mutations studied exhibited increased binding to CFB and formation of the C3 convertase.

*Department of Pediatrics, Clinical Sciences Lund, Lund University, Lund 22184, Sweden; †Division of Infection Medicine, Clinical Sciences Lund, Lund University, Lund 22184, Sweden; ‡Department of Laboratory Medicine, Center for Molecular Pathology, Lund University, Malmo¨ 20502, Sweden; xDepartment of Pathology, Ska˚ne University Hospital, Malmo¨ 20502, Sweden; {Department of Nephrology and Transplantation, Ska˚ne University Hospital, Malmo¨ 20502, Sweden; ‖Assistance Publique-Hopitaux de Paris, Hopital Europeén Georges-Pompidou, Service d’Immunologie Biologique, Paris 75015, France; #School of Natural Sciences, Linnæus University, Kalmar 39182, Sweden; and **Section of Clinical Immunology, Department of Immunology, Genetics and Pathology, Rudbeck Laboratory C5, Uppsala University, Uppsala 75185, Sweden

supported by grants from Agence Nationale de la Recherche (ANR Genopath 2009– 2012 09geno03101I and ANR Biotheque 2008–2011 R08086DS), Assistance Publique-Hoˆpitaux de Paris (Programme Hospitalier de Recherche Clinique AOM05130/ P051065 and AOM08198), and AIRG France (to V.F.-B.). K.N.-E. received faculty grants from Linnæus University, Kalmar. A preliminary version of the manuscript appeared in the Ph.D. thesis of Dr. Lisa Sartz.

1

L.S. and A.I.O. contributed equally to this work.

Received for publication February 1, 2011. Accepted for publication December 5, 2011. This work was supported by grants from the Swedish Research Council (K201065X-14008-10-3 to D.K. and 2009-4675 to K.N.-E.), the Torsten So¨derberg Foundation, the Fund for Renal Research, Crown Princess Lovisa’s Society for Child Care, Konung Gustaf V:s 80-a˚rsfond, and the Fanny Ekdahl Foundation (all to D.K.). D.K. is the recipient of a clinical-experimental research fellowship from the Royal Swedish Academy of Sciences. This work was also supported by the Queen Silvia Jubilee Fund (to L.S.), the Swedish Kidney Foundation and the Ska˚ne University Hospital Malmo¨ Fund for Medical Research (to M.E.J.). This work was also www.jimmunol.org/cgi/doi/10.4049/jimmunol.1100319

This study was presented as an oral presentation at Current Diagnosis and Therapy of Hemolytic Uremic Syndrome (HUS), May 18–20, 2009, Innsbruck, Austria, and at the Fourth International Workshop on Thrombotic Microangiopathies, October 1–3, 2009, Weimar, Germany. Address correspondence and reprint requests to Prof. Diana Karpman, Department of Pediatrics, Clinical Sciences Lund, Lund University, Lasarettsgatan 48, Lund 22185, Sweden. E-mail address: [email protected] The online version of this article contains supplemental material. Abbreviations used in this article: aHUS, atypical hemolytic uremic syndrome; C3b and C3c, cleaved C3; CFB, factor B; CFD, factor D; CFH, factor H; CFI, factor I; DAF, decay-accelerating factor; HUS, hemolytic uremic syndrome; MCP, membrane-cofactor protein; MPGN, membranoproliferative glomerulonephritis; RU, response unit. Copyright Ó 2012 by The American Association of Immunologists, Inc. 0022-1767/12/$16.00


2031

In addition to aHUS, membranoproliferative glomerulonephritis (MPGN) is associated with aberrations of the alternative pathway of complement, often exhibiting low C3 levels (12). MPGN may present with hematuria, nephrotic-range proteinuria, hypertension, and impaired renal function (13). MPGN is subclassified based on the localization of immune deposits consisting of IgG and/or C3 (12). MPGN types I and III are considered immunecomplex diseases, whereas type II, also known as dense deposit disease, is associated with complement activation via the alternative pathway and C3 deposition within the glomerular basement membrane (13). Recently, a heterozygous C3 mutation (923DDG) was described in familial dense deposit disease. The mutated C3 exhibited increased resistance to cleavage by C3 convertase; however, once cleaved, the active convertase generated by C3b923DDG exhibited increased resistance to decay by CFH, and C3b923DDG was resistant to inactivation by CFI in the presence of CFH (14). In this article, we describe the finding of a novel heterozygous gain-of-function C3 mutation, V1636A, leading to increased C3 binding to CFB and increased formation of the C3 convertase. The mutation, identified in a family with late-onset aHUS, could explain uncontrolled complement activation via the alternative pathway.

Materials and Methods Subjects One Swedish kindred with three affected individuals was investigated in this study (Patients 1–3). The family pedigree is presented in Fig. 1. All patients presented with clinical manifestations of HUS (nonimmune hemolytic anemia, thrombocytopenia, and acute renal failure) with simultaneous hypertension, hematuria, and nephrotic-range proteinuria. Clinical and histopathological findings in these patients are summarized in Tables I and II. Levels of C3, C4, CFB, CFH, CFI, and MCP are shown in Table III. Biopsies were taken during the acute phase of disease in Patients 1 and 2. In Patient 1, a biopsy was also taken after transplantation because of recurrence of aHUS. Patient 2 was on dialysis at the time of biopsy. Patient 3 underwent biopsy 8 mo after the onset of symptoms when acute signs of aHUS were no longer evident but chronic glomerular affection was present (hypertension, proteinuria, and renal failure). DNA extracts were obtained from Patients 1–3 and their family members. DNA extracts from an apparently healthy cohort of adult blood donors (n = 102) were screened as controls. Sera from Patients 2 and 3, as well as two healthy individuals, were used to purify C3b. Sera from the patients and five healthy controls were used for measurement of C3a. The study was performed with the approval of the Ethics Committee of the Medical Faculty at Lund University (protocol nos. 731-04 and 323-06) and with the written informed consent of all patients, family members, the wife of Patient 1, and healthy controls.

Mutation screening DNA was extracted from whole blood using the QIAamp DNA Blood Mini Kit (Qiagen, Hilden, Germany), as previously described (15), and from liver tissue (Patient 1) using the QIAamp DNA FFPE tissue kit (Qiagen). Exons of CFH (2), CFI (3), MCP (4), CFB (8), and C3 (primers listed in Supplemental Table I) were bidirectionally sequenced using the Big dye terminator kit (Applied Biosystems, Foster City, CA) and analyzed on an Applied Biosystems DNA Analyzer (model 3730).

CFH autoantibody assay Analysis of serum Abs to CFH was performed by ELISA, as previously described (16).

Blood samples, washed platelets, and serum Whole blood from patients and controls was drawn by venipuncture. Platelet-rich plasma (prepared from tubes containing 0.109 M sodium citrate [Becton Dickinson, Plymouth, U.K.]), washed platelets (prepared from tubes containing sodium citrate or K2EDTA [1.8 mg/ml; Becton Dickinson]), and serum were obtained, as previously described (17). Serum samples were stored at 280˚C until used.

C3b preparation C3 was purified from serum, according to a previously described method (18), and degraded to C3b by trypsin (Sigma-Aldrich, St. Louis, MO). C3 (1 mg) was incubated with trypsin (13 mg) in 1 mM HCl with 100 mM iodoacetamide (Sigma-Aldrich) in a total volume of 1 ml for 10 min at room temperature, followed by addition of soybean trypsin inhibitor (65 mg; Sigma-Aldrich) on ice, to stop the reaction. The resulting C3a and C3b fragments were separated by gel filtration on a Superose 6 column (GE Healthcare, Uppsala, Sweden). Gel filtration was performed in PBS (Medicago, Uppsala, Sweden) at 0.16 ml/min. Fractions of 0.5 ml were collected, and absorbance was read at 280 nm. Fractions were analyzed by C3 ELISA (19), pooled, and concentrated. Samples were kept at 280˚C until analyzed.

Mutant constructs A plasmid containing C3 cDNA in the psv expression vector was kindly provided by David Isenman (University of Toronto, Toronto, ON, Canada) (20). Mutant constructs were prepared using the QuikChange XL Site-directed mutagenesis kit (Stratagene, La Jolla, CA). Primers were forward, 59-CTTCACCGAGAGCATGGTTGCCTTTGGGTGCCCCAACTGAC-39 and reverse, 59-CTTCACCGAGAGCATGGTTGCCTTTGGGTGCCCCAACTGAC-39. The mutated fragment was sequenced, digested, and repositioned in the original vector by enzymatic digestion with FastDigest restriction enzymes XhoI/Bsp1407I (Fermentas Life Sciences, Helsingborg, Sweden).

Transient transfection COS-7 cells were cultured in DMEM (Invitrogen, Karlsruhe, Germany), supplemented with 100 U/ml penicillin, 100 mg/ml streptomycin, and 10% FBS (all from Invitrogen) to ∼95% confluence, followed by transient transfection with Lipofectamine (Invitrogen), according to the manu-

Table I. Clinical data related to patients investigated in this study Laboratory Findings Patient

1 2f 3f a

Sex

Male Male Male

Age at Diagnosis (y) a

52 48 42

Current Age (y)

Symptoms and Clinical Findings at Diagnosis b

Deceased 56 46

Edema, hypertension Dyspnea, edema, hypertension Jaundice, edema, hypertension, drusen

HA/DAT

Platelet Count

+/2 +/NA +/2

↓ ↓ ↓

Proteinuria/ Hematuria

Renal Failure c

End-stage, transplant End-stageg +h

d

+ /+ +d/+ +d/+

Relapses (n)

1e 0 3

Transient proteinuria at 24 y of age. Hepatitis C infection 1 y posttransplant. c Renal transplantation with graft donated by an unaffected brother 3 y after disease onset. d Nephrotic-range proteinuria. e Relapse of HUS 1 wk posttransplant. f ADAMTS13 activity was normal. g This patient developed end-stage renal disease at presentation and has not recovered renal function. h A protracted course of disease fluctuating for several months and partial recovery with decreased renal function. DAT, direct antiglobulin test; HA, hemolytic anemia determined by elevated lactate dehydrogenase, decreased haptoglobin, elevated reticulocyte count, and/or fragmented erythrocytes; NA, Not available; +, present; 2, negative. b

2032

GAIN-OF-FUNCTION C3 MUTATION IN aHUS

Table II. Pathological findings in patients Immunofluorescence/ Immunohistochemistry

Light Microscopic Glomerular Pathology

Patient

1

GBM Thickening and Double Contoursa

Mesangial Cell Proliferation

Lobulation

+ + + +

+ + + +

+ + + +

Pre-Tx Post-Tx

2 3

Microthrombi

Endothelial Cell Swelling

C3b

C1q

IgG

Electron Microscopy

+/2 + 2 +

+ + 2 +c

+ NA ++ ++

2 NA + 2

2 NA + 2

NA NA NA Endothelial detachment Podocyte foot process effacement No electron-dense deposits

a

Tram-tracks. Granular deposition along glomerular capillary walls. Mucoid intimal hyperplasia with fibrinoid necrosis was also noted. GBM, glomerular basement membrane; NA, not available; Post-Tx, posttransplantation; Pre-Tx, pretransplantation; +, Present; 2, absent; ++, present with excess labeling.

b c

facturer’s instructions. A total of 4 mg C3 DNA was added to each well. After 24 h, medium was changed to serum-free medium and cultured for an additional 48 h. Medium was removed, and a protease inhibitor mixture (complete Mini EDTA; Roche Diagnostic, Mannheim, Germany) was added. The supernatant was centrifuged to remove cells. C3 was quantified by ELISA, as described (6).

Healthcare) was used for washing and capping of unreacted carboxymethyl groups. Analyses were performed on a BIAcore 2000 instrument (GE Healthcare) at 25˚C in running buffer consisting of HBSS (PAA, Pashing, Austria), supplemented with 1 mM NiCl2 (Merck, Darmstadt, Germany). CFB was injected (up to 80 mg/ml in running buffer) at 35 ml/min over the flow cells with different C3b variants, and binding was assessed. In the second experimental setting, a polyclonal rabbit anti-human C3c Ab (Dako, Glostrup, Denmark) was first immobilized on flow cell surfaces. All surfaces were adjusted to bear equal amounts of Ab. This was followed by C3(H2O) injection (wild-type and mutant constructs diluted in running buffer) at 10 ml/min and a subsequent injection of CFB at 35 ml/min. C3 (H2O) contact times were varied to give similar responses of C3(H2O) capture by the Ab between variants. After subtraction of a control flow cell (treated as above but without addition of immobilized protein C3b in the first setting or anti-C3c in the second setting), sensorgrams were created using BIA Evaluation 4.1 software (GE Healthcare). From these sensorgrams, net binding (in RU) was calculated as the increment in RUs after CFB injection compared with baseline values corresponding to bound C3b in the first setting, as well as C3 (H2O) captured onto the Ab in the second setting. Formation of the C3 convertase and decay induced by decay-accelerating factor or CFH. C3 convertase was formed on the surface of a Biacore chip by serial injections of purified C3b alternating with CFB and factor D (CFD), as previously described (24). Convertase formation consisted of four serial injections of C3b, followed by CFB and CFD. The binding of CFB to the formed C3 convertase was tested by a subsequent injection of CFB at 35 ml/min, as described above. For decay assays, C3 convertases were formed, as described above, with the exception that the chip surface was obtained by immobilization of the polyclonal rabbit anti-human C3c Ab (as above), followed by serial injections of C3b (patients or control), CFB, and CFD. After the final C3b injection, levels of the convertase were adjusted to equivalent RUs to allow functional comparison. This was followed by injection (10 ml/min) of C3 variants (patients or control) over their respective C3 convertases. C3 was cleaved to C3b and deposited on the chip surface. A period of spontaneous convertase decay was followed by accelerated decay induced by injection of soluble decay-accelerating factor (DAF; a kind gift from professor Anna Blom, Division of Medical Protein Chemistry, Wallenberg Laboratory Malmo¨, Lund University) or CFH (Calbiochem, Darmstadt, Germany), both at 0.4 mM (10 ml/min), after which an additional period of time was

ADAMTS13 activity A modified collagen-binding assay was used to measure ADAMTS13 activity in plasma or serum, as previously described (21).

C3 functional assays C3 cleavage assay. C3 cleavage was measured in the fluid phase by crossed immunoelectrophoresis, as previously described (22). Patient serum was combined with normal serum, as a source of C3, for 30 min at 37˚C. The C3-cleavage reaction was terminated by dilution 1:3 in 0.075 M Diemal buffer (Barbitalum Natricum; APL, Gothenburg, Sweden) with EDTA (2 mM [pH 8.6]; Sigma). The resulting degree of C3 cleavage (measured as a decrease in C3 and an increase in cleaved C3 [C3c]) was estimated by the area under the curve, calculated by planimetry and analyzed by Digital board (GTCO; CalComp, Columbia, MD) as a quotient between the area outlined by immunoprecipitate (representing total C3) and the peak representing C3c. Results are presented as percentage C3 cleavage after subtraction of C3 cleavage in the control (normal and patient sera, which were first incubated separately for 30 min at 37˚C and then combined). C3 cleavage levels ,10% are considered normal (23). In certain samples, IgG was depleted from serum by passage through a Protein G Sepharose column (Amersham Biosciences, Uppsala, Sweden). C3a detection. C3a and C3a(desArg) were detected and quantified in sera using a C3a assay kit (Quidel, San Diego, CA), following the manufacturer’s instructions. Samples were stored at 280˚C until assayed. C3 binding to CFB determined by surface plasmon resonance. C3 binding to CFB (Complement Technology, Tyler, TX) was assayed using two experimental settings. In the first approach, purified C3b (diluted in 10 mM sodium acetate [pH 4]) was immobilized onto a CM5 Biacore sensorchip flow cell surface (GE Healthcare) via amine coupling. C3b preparations were immobilized corresponding to 1000 response units (RU) after Nhydroxysuccinimide/N-ethyl-N9-(dimethylaminopropyl) carbodiimide (GE Healthcare) activation of the CM5 matrix. Ethanolamine (1 M [ pH 8]; GE

Table III. Complement levels in Patients 1–3

Patient

1 2 3 a

C3 (770–1380 mg/l)a

C4 (120–330 mg/I)a

CFB (59–154%)a

CFH (69–154%)a

CFI (60–152%)a

MCP (500–1000 MFI)a,b

670 470 590

328 220 140

117 113 68

140 105 95

103 70 67

NA 719 NA

Normal range. Mean fluorescence intensity (MFI) analyzed by flow cytometry, as described (4). NA, Not analyzed. b


2033 C3 or C9 deposition on washed platelets was detected by incubation with chicken anti-human C3:FITC (1:2000; Diapensia, Linko¨ping, Sweden) or mouse anti-human C9 neoantigen (1:100; Hycult Biotechnology, Uden, The Netherlands) for 10 min. Chicken anti-human insulin:FITC (1:2000; Diapensia) and mouse IgG1 (1:100; Hycult Biotechnology) were used as the irrelevant Abs, respectively. Goat anti-mouse:FITC (1:700; Dako) was the secondary Ab. Acquisition and interpretation of flow cytometry. Platelets were analyzed using a BD FACSCantoII cytometer and FACSDiva software (Becton Dickinson Immunocytometry Systems, San Jose, CA), as previously described (17).

FIGURE 1. Pedigree of the investigated family, showing affected individuals (Patients 1–3) as filled symbols, and the results of the genetic investigation. Deceased patients are marked with a diagonal line. The mother of Patient 1 died undiagnosed during an episode of severe anemia and renal failure. The subscript under Patients 1–3 indicates the presence of a mutation in C3 and rare polymorphism in CFI, MCP, or CFH. Mutation in C3: V1636A. Rare polymorphisms were identified in CFI: IVS12+5, in MCP: A304V, and in CFH: Q950H. *This individual was tested and found to have the normal C3 (V1636) and CFI sequences. Patients 1 and 3 did not have the CFI polymorphism. allowed for spontaneous decay. Bound C3b was measured as the change from baseline (from the RU level at convertase formation) after cleavage and decay. Control flow cells were treated, as above, but without addition of anti-C3c at the first step of immobilization. Additional control cells were coated with anti-C3c onto which DAF or CFH was injected. C3b binding to CFH determined by surface plasmon resonance. Binding of CFH to C3b was assessed by injecting CFH at 0.4 mM and 10 ml/min. The experiment was performed on anti-C3c–coated CM5 surfaces bearing equal amounts of Ab. As a control, CFH was injected directly over the anti-C3c Ab. C3 and C9 deposition on platelets. C3 and C9 deposition on platelets from patients and controls was detected in platelet-rich plasma derived from EDTA-containing tubes (17). To determine whether mutated C3 induced complement deposition on the cell surface, washed human platelets (from tubes containing sodium citrate) obtained from donors with blood group O were used according to a previously described protocol (17). We previously showed that, under these conditions, heterologous combination of platelets from one donor with serum from another did not increase C3 and C9 deposition on platelets (17). Washed platelets (108/ml, final concentration in 100 ml) were incubated with an equal volume of patient (n = 2) or control (n = 5) serum for 2, 10, 20, or 30 min at 37˚C. The reaction was stopped by addition of 1:4 (v/v) ice-cold EDTA buffer [0.9 mM EDTA, 0.26 mM Na2HPO4(H2O), and 0.14 M NaCl (pH 7.2)], followed by centrifugation at 2000 3 g for 10 min. The supernatant separated from platelets in this experiment was stored at 280˚C until assayed for C3a, as described above. In a separate set of experiments, washed platelets were incubated with C3-depleted serum (Complement Technology), to which purified C3 from Patients 2 or 3 or one control, (final concentration 0.3 or 0.7 mg/ml) was added and incubated for 20 min, as above. C3-depleted serum was devoid of C3 and C3b, as assayed by immunoblotting (6) using anti-C3c Ab (1:1000; Dako). The C3-depleted serum contained 400 ng/ml C3a, corresponding to the level in normal serum; therefore, C3a was not assayed in the subsequent experiments.

Results C3 mutation A novel heterozygous missense C3 mutation V1636A was found in Patients 1–3 (described in Tables I–III). The mutation is located in exon 41, in the a-chain of C3 in one of three CFB-binding regions of C3b (25). This mutation was not found in 102 DNA samples from controls. The phenotypic consequences of this mutation were studied, as described below. In addition, rare polymorphisms were found in the MCP gene, A304V and CFI gene, IVS12+5G . T (Patient 2), and in the CFH gene, Q950H (Patient 3). The C3 mutation was the only mutation common to all affected family members. The mutation and polymorphisms are presented in Fig. 1 and Table IV. Functional assays C3 cleavage in patient sera. C3 cleavage was assayed in Patients 2 and 3. In serum from Patient 2, the C3-cleaving activity was elevated at 16%; in serum from Patient 3, the level was 20% (reference value , 10%). IgG was depleted from the serum of Patient 3, and the C3-cleaving activity was unaltered (20%), indicating that cleaving activity was not due to the presence of an Ab (such as C3 nephritic factor). Serum C3a levels. Patients 2 and 3 exhibited elevated levels of C3a in serum (1700 and 1600 ng/ml, respectively). Control sera (n = 5) had values of 350, 480, 510, 520, and 1000 ng/ml. Normal values of serum C3a are 71–589 ng/ml, according to the manufacturer of the C3a assay kit. C3 binding to CFB determined by surface plasmon resonance. Surface plasmon resonance analysis of purified C3b from Patients 2 and 3, as well as two healthy controls, showed increased binding of patient C3b to CFB in comparison with normal C3b (Fig. 2A). C3 convertase formation. The C3 convertase was formed on a Biacore sensor chip by serial injections of purified C3b from Patients 2 or 3 or from one control. This was followed by alternating injections of CFB and CFD. Formation of the C3 convertase is shown in Fig. 2B. After formation of the C3 convertase, CFB was injected, showing increased binding to the convertase formed with patient C3b compared with control C3b (Fig. 2B, inset).

Table IV. Molecular characteristics of genetic alterations

Patient

1 2a

3a

Complement Protein

Mutation or Polymorphism

Codon

Protein

Phenotype

Ref.

C3 C3 CFI

V1636A V1636A IVS12+5b

4973 T . C 4973 T . C IVS12+5 G . T

Val1636Ala Val1636Ala Intronic

Increased affinity for CFB and C3 convertase

This study

MCP

A304V

1058 C . T

Ala304Val

C3 CFH

V1636A Q950Hb

4973 T . C 2850 G . T

No documented abnormality and normal CFI serum levels Increased activation of alternative pathway on cell surface

(26)

b

Val1636Ala Gln950His

No documented abnormality

(28, 29)

All genetic alterations were heterozygous. a Tested and found not to have serum anti-CFH Abs. b These genetic alterations have also been detected in the healthy population, indicating they are rare polymorphisms.

(27)

2034


FIGURE 2. Binding of C3 to CFB. A, C3b binding to CFB. Purified C3b from Patients 2 and 3 and two controls was separately immobilized onto different flow cells of a CM5 chip. CFB was injected over the surfaces, and binding curves were visualized after subtraction of a blank flow cell. CFB binding to C3b from Patients 2 and 3 was higher than to C3b from controls. B, The C3 convertase was formed on Biacore CM5 chip flow cells by serial injections of purified C3b (from Patient 2), followed by CFB + CFD, as depicted in the main part of the panel. Convertase formation using C3b from Patient 3 and the control are not shown. After C3 convertase formation, CFB was injected, and a comparison was made between binding to the C3 convertase formed with C3b from Patient 2 or Patient 3 and from one control (inset). A higher response was detected when CFB was added to the C3 convertase formed with C3b from the patients compared with the control. C, CFB binding to C3(H2O) constructs after their immobilization to a BIAcore chip surface via Ab-affinity capture. An anti-human C3c Ab was first immobilized to different flow cells of a BIAcore chip in equal amounts (data not shown). C3(H2O) construct variants were captured by affinity corresponding to equal amounts (responses) in the different flow cells, after which CFB was injected, and binding was assessed. The different C3(H2O) variants were compared for CFB binding between experiments by an overlay of the CFB responses from the different C3(H2O) surfaces (lower boxed area) showing increased CFB binding to the V1636A mutant construct compared with WT construct V1636. The V1636A variant was composed of 50% wild-type V1636 and 50% mutant 1636A and, thus, was comparable to the heterozygous genotype of Patients 1–3.

Binding of C3 constructs to CFB. Wild-type and mutant C3 constructs were produced to study the V1636A mutation. The wildtype construct corresponded to V1636 (identical to the cDNA sequence). The heterozygote construct was composed of 50% wildtype V1636 and 50% mutant 1636A and was comparable with the C3 genotype of Patients 1–3. When CFB was injected over C3 (H2O) constructs bound to the C3c Ab-coated surface, a higher degree of binding of CFB to the C3 V1636A construct was detected compared with wild-type V1636 (Fig. 2C). Net formation of the C3 convertase after spontaneous and DAF- or CFH-induced decay. The C3 convertase was formed by serial injections of C3b from each patient or control, CFB, and CFD, followed by the respective C3, which deposited on the chip as

cleaved C3b. This was followed by a period of spontaneous decay (80 s), after which CFB displacement was induced by DAF or CFH. The results showed increased formation of the C3 convertase using C3b, followed by C3 from Patients 2 and 3, compared with the control. Increased convertase formation persisted, even after spontaneous and accelerated decay (Fig. 3). Measuring DRU from the start of accelerated decay for a period of 200 s showed that, after addition of DAF, the control curve decreased by 23 RU compared with 5 RU for Patient 2 and 3 RU for Patient 3. This pattern was not the same after addition of CFH: the control curve decreased by 11 RU, Patient 2 decreased by 24 RU, and Patient 3 decreased by 20 RU. These results indicated that the mutant C3 mediated excess formation of the C3 convertase but not excess


2035

FIGURE 3. Net formation of the C3 convertase after spontaneous and accelerated decay. Separate C3 convertases were formed by serial injections of C3b from Patient 2, Patient 3, or one control, as well as CFB and CFD over three different anti-C3c Ab BIAcore-CM5 surfaces. Formation of the C3 convertase occurred after ∼11,000 s, representing the baseline (adjusted as equivalent convertase formation). The corresponding C3 variant was injected over these convertases. Net C3b deposition was measured as a change in RU from baseline after a 80-s period of spontaneous convertase decay (buffer), followed by accelerated decay induced by DAF (A) or CFH (B). C, Control; Pat 2, Patient 2; Pat 3, Patient 3.

resistance to decay by DAF and CFH. Control cells (without addition of anti-C3c or with anti-C3c onto which DAF or CFH was injected) did not exhibit binding. To confirm that differences in decay were not due to altered binding affinity for CFH, equimolar amounts of CFH were injected over bound C3b from patients and controls; there was no difference in binding affinity (depicted as changes in RUs after addition of CFH; DRU in Fig. 4). No binding was detected when CFH was injected over the anti-C3c Ab-coated chip. Deposition of C3 and C9 on patient platelets. Washed platelets (from EDTA tubes) from Patient 2 and 3 displayed increased levels of C3 (24 and 16%, respectively) and C9 (15% and 13%, respectively) compared with washed platelets from healthy controls (C3: median 8%, range 6–11%; C9: median 7%, range 5–12%, n = 5). Serum from patients induced increased C3 and C9 deposition on washed platelets. Incubation of washed platelets with patient serum induced an increase in surface-bound C3 and C9 compared with control serum (Fig. 5). Similarly, incubation of washed platelets with C3-depleted serum, to which purified C3 from patient 2 or 3 was added, increased surface-bound C3 and C9 compared with purified C3 from the control (Fig. 6). Patient serum exposed to washed platelets induced C3a release. Levels of C3a were measured in the sera exposed to washed platelets. Patient sera induced an increase in C3a concentration over time (Fig. 7) compared with sera from healthy controls (n = 3).

FIGURE 4. Binding of purified C3b variants to CFH. The C3b-binding capacity of CFH was determined for the two patients and for a healthy control. There were no differences between the variants when equimolar amounts of CFH were injected. DRU, Differences in RUs before and after addition of CFH.

Discussion This study presents a novel C3 mutation, V1636A, in three family members with aHUS. The mutation is located in exon 41 in one of the three CFB-binding sites and induces enhanced C3 binding to CFB, thereby increasing formation of the C3 convertase. Increased formation persisted, even after spontaneous and accelerated decay induced by DAF and CFH. This mutation would be expected to enhance activation of the alternative pathway of complement and, indeed, all three patients exhibited clinical features of complement activation, such as decreased C3 levels in serum, increased C3a, increased C3 cleavage, and presence of C3 deposits on platelets and in glomerular capillary walls. Patient sera induced increased C3 and C9 deposition on normal washed human platelets over time and a parallel increase in C3a release. Previously described mutations in C3 exhibited loss-of-function with regard to the interaction between C3 and complement regulators, thus leading to activation of the alternative pathway. The V1636A would lead to a direct gainof-function with regard to formation of the C3 convertase. Serum from aHUS patients with CFH mutations localized at the C-terminal enabled increased C3 and C9 deposition on washed human platelets (17). The effect was specifically related to mutated CFH, as shown by combining washed platelets in CFHdepleted serum with purified CFH (normal or mutant variant). Mutant, but not normal, CFH allowed excess complement activation to occur on the cell surface. Addition of normal CFH to patient sera inhibited complement activation. In the current study,

2036


FIGURE 5. C3 and C9 binding to normal washed platelets exposed to serum over time. Normal washed platelets were incubated with serum from Patient 2 (d), Patient 3 (:), or controls (s). Patient sera induced increased C3 and C9 binding to normal platelets compared with normal sera. More C3 and C9 deposited over time.

excess complement activation on the cell surface was demonstrated using patient platelets, as well as normal washed platelets exposed to patient sera. However, the patients had other mutations/rare polymorphisms in CFH, CFI, or MCP, which also might have contributed to complement deposition on the platelet surface. C3-depleted serum was therefore used to specifically analyze whether the mutant variant of C3 induced complement activation on the platelet surface. The mutant variant of C3 resulted in excess C3 and C9 deposition on platelets. Experiments showing excess C3 and C9 deposition on the surface of washed platelets were carried out over a period of 30 min, indicating an increase in C3 and C9 deposition, as well as C3a release, over time. This suggested a prolonged half-life of the C3 convertase. To more specifically address this issue, binding experiments were carried out, showing that the mutant variant of C3 exhibited a higher affinity for CFB and more C3 convertase formation. The convertase formed from the mutant C3 variant persisted during spontaneous and DAF- or CFH-induced decay over time, indicating a prolonged half-life. DAF (41 kDa) or CFH (150 kDa) displace CFBb (60 kDa) bound to C3b in the C3bBb convertase. Surface plasmon resonance will not detect major changes in the curve when one molecule is replaced by another, particularly when the molecules have a similar molecular mass. A total decay of C3bBb was not visualized, rather there was a replacement of CFBb by DAF or CFH; therefore, a major change in the curve was not demonstrated. The results indicated that the mutant C3 most probably did not influence accelerated decay. Gain-of-function mutations affecting CFB have been described (8, 30). CFB mutations found in aHUS patients resulted in a highaffinity binding site to C3 and led to a hyperfunctional C3 convertase. Certain mutations led to resistance to decay by DAF

and/or CFH (8, 30). In addition, mutant CFB had the capacity to bind to iC3b, forming a C3 convertase from inactive C3b (30). Thus, mutations in both C3 and in CFB may lead to increased affinity between C3 and CFB. Interestingly, all three patients had normal levels of CFB. Increased formation of the C3 convertase with a prolonged half-life leads to more cleavage of C3 to C3b and consumption of C3, thus reducing serum levels. CFB levels do not necessarily decrease in parallel, even in patients with hypocomplementemia secondary to MPGN and C3 nephritic factor (23). Prolonged C3 convertase halflife, whether due to C3 nephritic factor or to the C3 mutant, V1636A, leads to excess C3 cleavage and, thus, lower C3, but not CFB, levels in serum. The family members described in this article exhibited a unique phenotype of aHUS. All affected individuals exhibited late onset of disease. In addition to the triad of HUS (DAT-negative hemolytic anemia, thrombocytopenia, and acute renal failure), these patients presented simultaneously with acute glomerulonephritis, as indicated by concurrent hematuria, hypertension, and nephroticrange proteinuria. In all three individuals, pathological findings suggested the presence of thrombotic microangiopathy, as well as membranoproliferative features, such as glomerular lobulation, mesangial proliferation, tram-tracking of the glomerular basement membrane, and C3 deposits in the glomeruli. All of these light microscopy features may be associated with chronic thrombotic microangiopathy and with MPGN (31). The similarity in clinical and pathological presentation in the three patients suggested that the common C3 mutation was an underlying cause of the mutual unique phenotype. We speculated that increased formation of the C3 convertase may give rise to a mixed phenotype combining features of aHUS and glomerulonephritis resembling MPGN, al-

FIGURE 6. Binding of C3 and C9 to washed platelets exposed to purified C3 added to C3-depleted serum. Normal washed platelets were incubated with C3-depleted serum, to which purified C3 from Patient 2 (d; 0.3 mg/ml final concentration), Patient 3 (:; 0.7 mg/ml final concentration), or one control (s; 0.3 mg/ml or ); 0.7 mg/ml final concentration) was added. Purified C3 from patients induced increased C3 and C9 binding compared with the control.

FIGURE 7. C3a release over time in sera exposed to washed platelets. C3a was measured in sera exposed to washed platelets over a period of 30 min. All sera C3a levels were adjusted to a baseline “0” level at the exposure to serum. Sera from Patient 2 (:) and Patient 3 (n) exhibited steadily increasing levels, whereas sera from three controls (d) showed persistently low levels.

The Journal of Immunology though ultramorphology in Patient 3 did not exhibit features of MPGN. In MPGN, the C3 nephritic factor stabilizes the C3 convertase, whereas in the patients described in this article, the C3 mutation had a similar effect, which may explain the phenotypic overlap resembling MPGN. In summary, this study presents a novel C3 mutation, leading to increased formation of the C3 convertase, in three related individuals affected with aHUS. This C3 mutation promotes uncontrolled complement activation, as demonstrated in the patients by decreased serum C3 levels and C3 deposits in renal tissue. To our knowledge, this is the first C3 mutation to be described leading to a direct gain-of-function, and we suggest that its functional consequences may lead to increased activation of the alternative pathway.

2037

12.

13.

14.

15.

16.

Acknowledgments We thank Prof. Martin Olsson (Department of Transfusion Medicine, Lund University) for control samples, Prof. Anna Blom (Division of Medical Protein Chemistry, Wallenberg Laboratory Malmo¨, Lund University) for soluble DAF, Ramesh Tati (Department of Pediatrics, Lund University) for help with figures, Kerstin Sandholm (School of Natural Sciences, Linnæus University, Kalmar) for purification of C3b, Dr. Lillemor Skattum and Prof. Lennart Truedsson (Department of Clinical Immunology, Ska˚ne University Hospital, Lund) for helpful discussions, and Dr. Sira Sooparb (Nephrology Clinic, Bumrungrad International Hospital, Bangkok, Thailand) for clinical information about Patient 3.

Disclosures K.N.-E. holds a patent related to the modification of biomaterials, V.F.-B. was consultant for Alexion Pharmaceuticals, and D.K. was the national coordinator in Sweden of the Eculizumab trial (Alexion Pharmaceuticals) in patients with aHUS.

References 1. Noris, M., and G. Remuzzi. 2009. Atypical hemolytic-uremic syndrome. N. Engl. J. Med. 361: 1676–1687. 2. Richards, A., M. R. Buddles, R. L. Donne, B. S. Kaplan, E. Kirk, M. C. Venning, C. L. Tielemans, J. A. Goodship, and T. H. Goodship. 2001. Factor H mutations in hemolytic uremic syndrome cluster in exons 18-20, a domain important for host cell recognition. Am. J. Hum. Genet. 68: 485–490. 3. Fremeaux-Bacchi, V., M. A. Dragon-Durey, J. Blouin, C. Vigneau, D. Kuypers, B. Boudailliez, C. Loirat, E. Rondeau, and W. H. Fridman. 2004. Complement factor I: a susceptibility gene for atypical haemolytic uraemic syndrome. J. Med. Genet. 41: e84. 4. Fremeaux-Bacchi, V., E. A. Moulton, D. Kavanagh, M. A. Dragon-Durey, J. Blouin, A. Caudy, N. Arzouk, R. Cleper, M. Francois, G. Guest, et al. 2006. Genetic and functional analyses of membrane cofactor protein (CD46) mutations in atypical hemolytic uremic syndrome. J. Am. Soc. Nephrol. 17: 2017–2025. 5. Zipfel, P. F., M. Edey, S. Heinen, M. Jo´zsi, H. Richter, J. Misselwitz, B. Hoppe, D. Routledge, L. Strain, A. E. Hughes, et al. 2007. Deletion of complement factor H-related genes CFHR1 and CFHR3 is associated with atypical hemolytic uremic syndrome. PLoS Genet. 3: e41. 6. Fre´meaux-Bacchi, V., E. C. Miller, M. K. Liszewski, L. Strain, J. Blouin, A. L. Brown, N. Moghal, B. S. Kaplan, R. A. Weiss, K. Lhotta, et al. 2008. Mutations in complement C3 predispose to development of atypical hemolytic uremic syndrome. Blood 112: 4948–4952. 7. Maga, T. K., C. J. Nishimura, A. E. Weaver, K. L. Frees, and R. J. Smith. 2010. Mutations in alternative pathway complement proteins in American patients with atypical hemolytic uremic syndrome. Hum. Mutat. 31: E1445–E1460. 8. Goicoechea de Jorge, E., C. L. Harris, J. Esparza-Gordillo, L. Carreras, E. A. Arranz, C. A. Garrido, M. Lo´pez-Trascasa, P. Sańchez-Corral, B. P. Morgan, and S. Rodrı´guez de Co´rdoba. 2007. Gain-of-function mutations in complement factor B are associated with atypical hemolytic uremic syndrome. Proc. Natl. Acad. Sci. USA 104: 240–245. 9. Delvaeye, M., M. Noris, A. De Vriese, C. T. Esmon, N. L. Esmon, G. Ferrell, J. Del-Favero, S. Plaisance, B. Claes, D. Lambrechts, et al. 2009. Thrombomodulin mutations in atypical hemolytic-uremic syndrome. N. Engl. J. Med. 361: 345–357. 10. Loirat, C., M. Noris, and V. Fremeaux-Bacchi. 2008. Complement and the atypical hemolytic uremic syndrome in children. Pediatr. Nephrol. 23: 1957–1972. 11. Moore, I., L. Strain, I. Pappworth, D. Kavanagh, P. N. Barlow, A. P. Herbert, C. Q. Schmidt, S. J. Staniforth, L. V. Holmes, R. Ward, et al. 2010. Association

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

of factor H autoantibodies with deletions of CFHR1, CFHR3, CFHR4, and with mutations in CFH, CFI, CD46, and C3 in patients with atypical hemolytic uremic syndrome. Blood 115: 379–387. Benz, K., and K. Amann. 2009. Pathological aspects of membranoproliferative glomerulonephritis (MPGN) and haemolytic uraemic syndrome (HUS)/ thrombocytic thrombopenic purpura (TTP). Thromb. Haemost. 101: 265–270. Appel, G. B., H. T. Cook, G. Hageman, J. C. Jennette, M. Kashgarian, M. Kirschfink, J. D. Lambris, L. Lanning, H. U. Lutz, S. Meri, et al. 2005. Membranoproliferative glomerulonephritis type II (dense deposit disease): an update. J. Am. Soc. Nephrol. 16: 1392–1403. Martıńez-Barricarte, R., M. Heurich, F. Valdes-Can˜edo, E. Vazquez-Martul, E. Torreira, T. Montes, A. Tortajada, S. Pinto, M. Lopez-Trascasa, B. P. Morgan, et al. 2010. Human C3 mutation reveals a mechanism of dense deposit disease pathogenesis and provides insights into complement activation and regulation. J. Clin. Invest. 120: 3702–3712. Vaziri-Sani, F., L. Holmberg, A. G. Sjo¨holm, A. C. Kristoffersson, M. Manea, V. Fre´meaux-Bacchi, I. Fehrman-Ekholm, R. Raafat, and D. Karpman. 2006. Phenotypic expression of factor H mutations in patients with atypical hemolytic uremic syndrome. Kidney Int. 69: 981–988. Dragon-Durey, M. A., C. Loirat, S. Cloarec, M. A. Macher, J. Blouin, H. Nivet, L. Weiss, W. H. Fridman, and V. Fre´meaux-Bacchi. 2005. Anti-Factor H autoantibodies associated with atypical hemolytic uremic syndrome. J. Am. Soc. Nephrol. 16: 555–563. Sta˚hl, A. L., F. Vaziri-Sani, S. Heinen, A. C. Kristoffersson, K. H. Gydell, R. Raafat, A. Gutierrez, O. Beringer, P. F. Zipfel, and D. Karpman. 2008. Factor H dysfunction in patients with atypical hemolytic uremic syndrome contributes to complement deposition on platelets and their activation. Blood 111: 5307– 5315. Basta, M., and C. H. Hammer. 1991. A rapid FPLC method for purification of the third component of human and guinea pig complement. J. Immunol. Methods 142: 39–44. Henningsson, A. J., J. Ernerudh, K. Sandholm, S. A. Carlsson, H. Granlund, C. Jansson, D. Nyman, P. Forsberg, and K. Nilsson Ekdahl. 2007. Complement activation in Lyme neuroborreliosis—increased levels of C1q and C3a in cerebrospinal fluid indicate complement activation in the CNS. J. Neuroimmunol. 183: 200–207. Taniguchi-Sidle, A., and D. E. Isenman. 1992. Mutagenesis of the Arg-Gly-Asp triplet in human complement component C3 does not abolish binding of iC3b to the leukocyte integrin complement receptor type III (CR3, CD11b/CD18). J. Biol. Chem. 267: 635–643. Gerritsen, H. E., P. L. Turecek, H. P. Schwarz, B. La¨mmle, and M. Furlan. 1999. Assay of von Willebrand factor (vWF)-cleaving protease based on decreased collagen binding affinity of degraded vWF: a tool for the diagnosis of thrombotic thrombocytopenic purpura (TTP). Thromb. Haemost. 82: 1386–1389. Peters, D. K., A. Martin, A. Weinstein, J. S. Cameron, T. M. Barratt, C. S. Ogg, and P. J. Lachmann. 1972. Complement studies in membrano-proliferative glomerulonephritis. Clin. Exp. Immunol. 11: 311–320. Skattum, L., U. Ma˚rtensson, and A. G. Sjo¨holm. 1997. Hypocomplementaemia caused by C3 nephritic factors (C3 NeF): clinical findings and the coincidence of C3 NeF type II with anti-C1q autoantibodies. J. Intern. Med. 242: 455–464. Jokiranta, T. S., J. Westin, U. R. Nilsson, B. Nilsson, J. Hellwage, S. Lo¨fa˚s, D. L. Gordon, K. N. Ekdahl, and S. Meri. 2001. Complement C3b interactions studied with surface plasmon resonance technique. Int. Immunopharmacol. 1: 495–506. Janssen, B. J. C., A. Christodoulidou, A. McCarthy, J. D. Lambris, and P. Gros. 2006. Structure of C3b reveals conformational changes that underlie complement activity. Nature 444: 213–216. Caprioli, J., M. Noris, S. Brioschi, G. Pianetti, F. Castelletti, P. Bettinaglio, C. Mele, E. Bresin, L. Cassis, S. Gamba, et al; International Registry of Recurrent and Familial HUS/TTP. 2006. Genetics of HUS: the impact of MCP, CFH, and IF mutations on clinical presentation, response to treatment, and outcome. Blood 108: 1267–1279. Fang, C. J., V. Fremeaux-Bacchi, M. K. Liszewski, G. Pianetti, M. Noris, T. H. Goodship, and J. P. Atkinson. 2008. Membrane cofactor protein mutations in atypical hemolytic uremic syndrome (aHUS), fatal Stx-HUS, C3 glomerulonephritis, and the HELLP syndrome. Blood 111: 624–632. Caprioli, J., F. Castelletti, S. Bucchioni, P. Bettinaglio, E. Bresin, G. Pianetti, S. Gamba, S. Brioschi, E. Daina, G. Remuzzi, and M. Noris; International Registry of Recurrent and Familial HUS/TTP. 2003. Complement factor H mutations and gene polymorphisms in haemolytic uraemic syndrome: the C257T, the A2089G and the G2881T polymorphisms are strongly associated with the disease. Hum. Mol. Genet. 12: 3385–3395. Neumann, H. P., M. Salzmann, B. Bohnert-Iwan, T. Mannuelian, C. Skerka, D. Lenk, B. U. Bender, M. Cybulla, P. Riegler, A. Ko¨nigsrainer, et al. 2003. Haemolytic uraemic syndrome and mutations of the factor H gene: a registrybased study of German speaking countries. J. Med. Genet. 40: 676–681. Roumenina, L. T., M. Jablonski, C. Hue, J. Blouin, J. D. Dimitrov, M. A. Dragon-Durey, M. Cayla, W. H. Fridman, M. A. Macher, D. Ribes, et al. 2009. Hyperfunctional C3 convertase leads to complement deposition on endothelial cells and contributes to atypical hemolytic uremic syndrome. Blood 114: 2837–2845. Pickering, M. C., and H. T. Cook. 2008. Translational mini-review series on complement factor H: renal diseases associated with complement factor H: novel insights from humans and animals. Clin. Exp. Immunol. 151: 210–230.