Lack of association between polymorphisms in C4b-binding protein and atypical haemolytic uraemic syndrome in the Spanish population

Abstract

Dysregulation of the alternative pathway of complement activation, caused by mutations or polymorphisms in the genes encoding factor H, membrane co-factor protein, factor I or factor B, is associated strongly with predisposition to atypical haemolytic uraemic syndrome (aHUS). C4b-binding protein (C4BP), a major regulator of the classical pathway of complement activation, also has capacity to regulate the alternative pathway. Interestingly, the C4BP polymorphism p.Arg240His has been associated recently with predisposition to aHUS and the risk allele His240 showed decreased capacity to regulate the alternative pathway. Identification of novel aHUS predisposition factors has important implications for diagnosis and treatment in a significant number of aHUS patients; thus, we sought to replicate these association studies in an independent cohort of aHUS patients. In this study we show that the C4BP His240 allele corresponds to the C4BP*2 allele identified previously by isoelectric focusing in heterozygosis in 1·9–3·7% of unrelated Caucasians. Crucially, we found no differences between 102 unrelated Spanish aHUS patients and 128 healthy age-matched Spanish controls for the frequency of carriers of the His240 C4BP allele. This did not support an association between the p.Arg240His C4BP polymorphism and predisposition to aHUS in the Spanish population. In a similar study, we also failed to sustain an association between C4BP polymorphisms and predisposition to age-related macular degeneration, another disorder which is associated strongly with polymorphisms in factor H, and is thought to involve alternative pathway dysregulation.

Introduction

Human C4b-binding protein (C4BP) is an abundant plasma protein, synthesized primarily in the liver, that presents three isoforms, α7β1, α7β0 and α6β1 [1,2]. In normal plasma C4BPα7β1 is the major isoform. It is composed of seven identical α-chains (70 kDa) and one β-chain (45 kDa), linked covalently by their C-terminal regions [3,4]. The C4BPA and C4BPB genes are located within the regulator of complement activation gene cluster in chromosome 1q32 [5,6]. They are arranged in tandem with the 3′ end of the C4BPB gene located 4172 base pairs upstream of the 5′ end of the C4BPA gene [6]. C4BP is the principal fluid phase regulator of the classical pathway of complement activation [7–9]. C4BP also acts as a co-factor of factor I in the proteolytic inactivation of C3b and therefore also has potential to regulate the alternative pathway of complement activation [10]. C4BP is, however, a much weaker regulator than factor H when C3b is degraded on surfaces. Each α-chain of C4BP contains a C4b/C3b binding site, but only up to four molecules of C4b or C3b can bind one C4BP molecule [11]. The binding site for C4b spans Short Consensus Repeats (SCRs) 1–3 [12,13] and that for C3b also extends to SCR4 [10]. The α-chain of C4BP is a polymorphic protein in humans with three alleles identified by isoelectric focusing (IEF), one common (C4BP*1) and two relatively uncommon (C4BP*2 and C4BP*3) [14,15].

Haemolytic uraemic syndrome (HUS) is characterized by thrombocytopenia, Coomb's test negative microangiopathic haemolytic anaemia and acute renal failure. Endothelial cell injury appears to be the primary event in the pathogenesis of HUS. Endothelial damage triggers a cascade of events that results in the formation of platelet-fibrin hyaline microthrombi that occlude arterioles and capillaries. A hallmark of HUS is the presence of schistocytes (fragmented cells) that generate as the red blood cells traverse these partially occluded microvessels [16]. The typical form of HUS follows a diarrhoeal prodrome and is associated with 0157:H7 Escherichia coli infections. However, 5–10% of HUS patients lack an association with this type of E. coli infections. This atypical form of HUS (aHUS) occurs in both adults and in very young children and has the poorest long-term prognosis. aHUS is associated with mutations or polymorphisms in the genes encoding the complement regulatory proteins factor H (CFH) [17–23], membrane co-factor protein (MCP) [24–27] and with mutations in the complement activating components factor B (CFB) [28] and C3 genes [29]. Importantly, mutations in the complement regulators factor H, MCP and factor I are loss-of-function mutations [27,30,31], while mutations in the complement activator factor B are gain-of-function mutations [28]. aHUS has also been associated with the deletion of the CFHR1 and CFHR3 genes [32] and the presence of antibodies against factor H [33,34].

The data available support that aHUS involves alternative complement pathway dysregulation and probably develops as a consequence of defective protection of cellular surfaces from complement activation due to an improper function of different complement proteins [35]. In addition, it is generally accepted that multiple hits, involving plasma and membrane-associated complement regulatory proteins, as well as complement activators, are probably required to cause dysregulation and significantly impair protection to host tissues [24,28,36].

The C4BP has been implicated in the regulation of the alternative pathway [10]. Recently, a case–control study has tested the possibility that genetic variations in the C4BPA and C4BPB genes were associated with predisposition to aHUS and identified a relatively uncommon C4BPA polymorphism, p.Arg240His, that was present in 6/166 of aHUS patients versus 5/542 normal controls (χ² = 6·021; P = 0·014) [37]. Interestingly, the aHUS-associated C4BP allele His240 showed decreased binding to C3b compared with the C4BP Arg240 allele.

In this study, we have established that the p.Arg240His is the nucleotide substitution that originates the C4BP*2 protein polymorphic variant described previously by IEF and found to be present in heterozygosis in 1·9–3·7% of unrelated Caucasians. Most importantly, we show that this polymorphism is not associated with aHUS or age-related macular degeneration (AMD) in the Spanish population.

Materials and methods

Patients and controls

Our series of aHUS patients include 125 unrelated individuals. aHUS was diagnosed because of the presence of one or more episodes of microangiopathic haemolytic anaemia and thrombocytopenia defined on the basis of haematocrit < 30%, haemoglobin < 10 mg/dl, serum lactate dehydrogenase > 460 U/l, undetectable haptoglobin, fragmented erythrocytes in the peripheral blood smear and platelet count < 150 000/µl, associated with acute renal failure. Patients with Stx-HUS, defined as the presence of Shiga toxin in the stools (by the Vero cell assay) and/or of serum antibodies against Shiga toxin [by enzyme-linked immunosorbent assay (ELISA)] and/or lipopolysaccharide (O157, O26, O103, O111 and O145, by ELISA) were excluded. None of the aHUS patients had thrombotic thrombocytopenic purpura-like manifestations. Age at onset ranges from 1 month to 70 years, with 57 patients developing the disease aged under 13 years and 68 patients aged over 21 years. All patients were analysed routinely for mutations in the CFH, MCP, CFI, C3 and CFB genes, tested for the presence of anti-factor H autoantibodies, genotyped for high-risk polymorphisms in the CFH and MCP genes and analysed for copy number variation of CFHR1 and CFHR3and rearrangements in the CFH-CFHR1-5 gene region. The overall number of patients who carry mutations or rearrangements in the complement genes or have detectable anti-factor H autoantibodies represent an estimable 50% and 6% of all the atypical HUS patients in our series, respectivelly.

For the studies reported here, the 125 unrelated aHUS patients in our cohort were divided into two groups, one including 102 unrelated Spaniards living in Spain and another group of 23 aHUS patients who were referred to us from outside Spain. This last group includes seven patients from other European countries, six from the USA, six from South America and four from Tunisia. Our control population for the aHUS cohort includes 128 healthy age-matched unrelated Spanish individuals from similar geographical regions to the aHUS patients.

The C4BP studies presented here also included the analysis of 143 patients > 60 years old with AMD, who presented with advanced choroidal neovascularization and drusen in both eyes, and the corresponding control cohort of 117 eye-examined, age-matched healthy Spanish controls with no family history of AMD.

Genomic DNA was generated from a buccal swab (BuccalAmp DNA Extraction kit; Epicentre Biotechnologies, Madison, WI, USA) or peripheral blood leucocytes using standard procedures. All protocols included in these studies have been approved by national and/or local institutional review boards, and all subjects gave their informed consent.

In addition, we had DNA samples available for these studies from all members of five of the original pedigrees that were used to demonstrate the IEF polymorphism in the α-chain of C4BP [15]. None of these individuals were included in the case–control association studies presented here.

The IEF analysis of C4BP

The IEF analysis of C4BP was performed as described [14] by using fresh ethylenediamine tetraacetic acid-containing serum or serum samples stored at –80°C. Samples were treated with neuraminidase before immunoprecipitation with a rabbit polyclonal antibody against human C4BP. Immunoprecipitates were analysed by IEF under completely denaturing conditions on vertical 4·5% polyacrylamide slab gels and stained with Coomassie blue for analysis.

C4BPA sequencing and genotyping analysis

All C4BPA exons were amplified from genomic DNA using the primers shown in Table 1. These amplifications were performed with the Certamp Kit for Complex Amplifications (purchased from Biotools, Madrid, Spain). The thermal cycle started with 3 min at 94°C followed by 35 cycles of 30 s at 94°C for denaturation, 40 s at the annealing temperature shown in Table 1 and another 40 s at 72°C. These 35 cycles were followed by a final elongation step of 6 min at 72°C. Automatic DNA sequencing was performed in an ABI 3730 sequencer using a dye terminator cycle sequencing kit (Applied Biosystems, NJ, USA) with one of the two primers used in the polymerase chain reaction (PCR) amplification of the exons. Genotyping of the polymorphism c.719G>A (Arg240His) in exon 7 and c.1615G>C (Glu539Gln) in exon 11 of C4BP was also performed by automatic DNA sequencing of PCR-derived amplicons.

Table 1.

Primers for the amplification and sequencing of the C4BPA exons.

Name	Primers 5′-3′	Temp °C
C4BPA-E2F	CAGAGGCCAATCCTTACTG	58
C4BPA-E2R	AGTGAATGCACAGATGATCTTC
C4BPA-E3F	ATCACTCACTAGGTGTATTGAACAG	58
C4BPA-E3R	ATGTACATTGAAATCAGTAATTGAC
C4BPA-E4F	TGCTGCAATTATTTAACTGGTC	58
C4BPA-E4R	AAATGAGAATTCTCAACGTCTATC
C4BPA-E5/6F	TACTGAATGGAGTGCTTAATGG	58
C4BPA-E5/6R	ATTGCATCAGAGCGATAGTTAG
C4BPA-E7F	CCTTGTATCTACTTGACATGCAC	60
C4BPA-E7R	CTGTAATTCATAGTAGGCCTAT
C4BPA-E8F	GCACATTTGAGCGAATGTC	58
C4BPA-E8R	TTCTGGCACTAACAGTTTACTAATC
C4BPA-E9F	AGCAGACTTGGAGTGTCTTG	58
C4BPA-E9R	CAAGAATCATGCAAGGTATATG
C4BPA-E10F	TTGGCTACGTGCTCTTACAG	56
C4BPA-E10R	GAGCCTATATTCATACTAGAGCAAG
C4BPA-E11F	GTGTCTTAATGTATCTCAGCTGTC	60
C4BPA-E11R	AATCCTGACAAGTCAGATTTCAC
C4BPA-E12F	GGTCTTGGTTCAATCCTGTAG	56
C4BPA-E12R	CCAAGTACCTATTATTTGTGGC

Statistical analysis

To acquire statistical data comparable with that generated by Blom et al. [34], we analysed the differences in the C4BP allelic frequencies between the aHUS, AMD and control groups by the Pearson's χ² test. A P-value of 0·05 or less was considered to be statistically significant.

Results

Identification of the amino acid substitutions that characterize the three IEF C4BPA alleles

Genetic polymorphism of C4BPA was first described at the protein level using IEF [15]. Subsequently, in a sample of 516 normal unrelated Caucasians, three C4BPA alleles, denominated C4BP*1, C4BP*2 and C4BP*3, were identified with estimated allele frequencies of 0·986, 0·010 and 0·004 respectively [15]. C4BP*2 and C4BP*3 have been found only in heterozygosis with C4BP*1.

To identify the amino acid substitutions that characterize the C4BPA IEF alleles we selected one C4BP*1/C4BP*1 homozygote, one C4BP*1/C4BP*2 heterozygote and one C4BP*1/C4BP*3 heterozygote and determined the nucleotide sequences of all the coding exons of C4BPA in their genomic DNA. Figure 1 summarizes these analyses. c.719G>A was the only nucleotide change resulting in an amino acid sustitution (Arg240His) observed between the C4BP*1/C4BP*1 homozygote and the C4BP*1/C4BP*2 heterozygote. Similarly, c.1615G>C underlies the single amino acid difference (Glu539Gln) between the C4BP*1/C4BP*1 homozygote and the C4BP*1/C4BP*3 heterozygote.

Fig. 1.

Sequence analysis of the C4BPA isoelectric focusing (IEF) alleles. IEF on polyacrylamide gels under completely denaturing conditions of neuraminidase-treated affinity-purified C4b-binding protein (C4BP) proteins from C4BP*1/C4BP*3, C4BP*1/C4BP*1 and C4BP*1/C4BP*2 individuals is shown on the left side of the figure. Each C4BPα polypeptide chain discloses a banding pattern composed of three bands, the cathodal band being the more heavily stained. The cathodal bands corresponding to the three C4BPA alleles are indicated by arrows. On the right side fragments of the electropherograms corresponding to the regions of exon 7 (SCR4) and exon 11 (SCR8) of C4BPA gene from the same C4BP*1/C4BP*3, C4BP*1/C4BP*1 and C4BP*1/C4BP*2 individuals are shown to illustrate the nucleotide substitutions that identify the C4BP*2 (c.719G>A; Arg240His) and C4BP*3 (c.1615G>C; Glu539Gln) alleles.

To confirm that the Arg240His and the Glu539Gln amino acid changes identified the C4BP*2 and C4BP*3 C4BPA allelic variants, we genotyped all members of five available pedigrees in which we have previously shown segregation of the C4BP*2 and C4BP*3 alleles (Fig. 2). In agreement with the sequencing results, data from these families showed co-segregation of His240 with C4BP*2 alleles and Gln539 with C4BP*3, demonstrating that the amino acid substitutions Arg240His in SCR4 and Glu539Gln in SCR8 identify the C4BP*2 and C4BP*3 allelic variants respectively.

Fig. 2.

Segregation analysis of the C4BPA alleles. Five pedigrees carrying the C4BP*2 or the C4BP*3 isoelectric focusing (IEF) protein variants are shown to illustrate the perfect segregation of these alleles with the c.719G>A (His240; C4BP*2) and c.1615G>C (Gln539; C4BP*3) nucleotide substitutions. The numbers on the right of the pedigrees are the codes for the samples.

C4BPA genotyping analysis in the aHUS and AMD Spanish cohorts

Recently, Blom et al. [34] reported that the p.Arg240His polymorphism is associated with aHUS and performed functional analyses showing that a recombinant C4BP molecule composed of His240 C4BPα polypeptide chains had impaired capacity to regulate the alternative pathway when compared with a similar molecule composed of Arg240 C4BPα polypeptide chains. Both binding of C4BP to C3b on surfaces and in the fluid phase was decreased in the His240 C4BP molecules, resulting in impaired co-factor capacity for degradation of C3b by factor I. In contrast, C4b-binding and degradation of C4b were not affected by this C4BPA polymorphism. These are important data that fit with current hypotheses regarding the pathogenesis of aHUS and provide a novel candidate gene contributing to aHUS predisposition. However, the aHUS cohort used by Blom et al. [34] to calculate the frequency of the His240 C4BPA allele included related patients carrying the Arg240 C4BPA allele, leading to an overestimation of the frequency of the His240 C4BPA allele. In addition, the replication study performed by the same authors in an independent cohort of aHUS patients was at the limit of statistical significance (χ² = 3·808; P = 0·051). To resolve these uncertainties and clarify the association of this polymorphism with disease, we set out to test the association of the p.Arg240His C4BP polymorphism with aHUS in our independent cohort of aHUS patients. In addition, we extended these association studies to our cohort of AMD patients as AMD also involves dysregulation of the alternative pathway [38–40]. Our cohorts of patients consist of 102 unrelated aHUS Spanish patients and 143 unrelated AMD Spanish patients (see Materials and methods). The results of these studies are summarized in Table 2 and illustrate that there are no differences in the frequencies of the C4BPA His240 carriers between aHUS patients or AMD patients and their respective control populations. Therefore, we concluded that the p.Arg240His C4BPA polymorphism is not associated with increased risk to aHUS or to AMD in the Spanish population. None of the 23 non-Spanish patients in our cohort of aHUS patients carried the His240 C4BPA allele. Similarly, none of the 17 patients in the Spanish membraneproliferative glomerulonephritis type II cohort carried the His240 C4BPA allele, but this number of patients is too small to draw any conclusion.

Table 2.

Frequencies of C4BPA polymorphisms in patients and controls

Discussion

The search for novel polymorphisms conferring increased risk for aHUS is important, as current genetic analysis of the CFH, MCP, CFI, CFB and C3genes identifies only mutations in approximately 50% of unrelated patients. Novel genes harbouring aHUS-associated polymorphisms that are supported fully by statistical and biological significance are excellent candidate genes for also carrying mutations associated with aHUS; these genes should be added to those analysed routinely in aHUS patients. Based on the studies reported by [37], C4BPA may be one of these novel genes. However, the association studies implicating C4BPA with aHUS were not supported fully by statistical significance in a replication study by the same authors. To clarify these data, we sought to replicate the association of the p.Arg240His C4BPA polymorphism with aHUS in our cohort of patients. Unfortunately, however, our results found no differences between patients and controls for the frequency of carriers of the C4BPA His240 allele. One possible explanation for the differences between our results and those reported by Blom et al. [34] is that the frequency of this relatively uncommon C4BPA polymorphism varies between different populations (Table 2). This may be particularly critical in the case of aHUS, with limited numbers of patients from many different geographical origins. Since 1983 we have analysed three different Caucasians populations, two from the New York area in the United States and one from Spain, with frequencies of His240 carriers ranging from 0·78 to 3·7. Notably, the frequencies of His240 carriers in the two aHUS cohorts and the control populations described by Blom et al. [34] are included within this range (Table 2). In the studies presented in this report we selected our patient and control groups to include only Spaniards from similar geographical areas and found identical numbers of carriers in two independent control populations and two cohorts of patients, one with aHUS and the other with AMD (Table 2).

In summary, our data do not support an association between the p.Arg240His C4BPA polymorphism and aHUS. Therefore, despite functional analyses which potentially give biological significance for the association of this polymorphism with aHUS, the statistics supporting the association are insufficient to consider the inclusion of the C4BPA gene in the routine mutational analysis of complement genes in aHUS patients.