Article Text

Download PDFPDF

The high frequency of complement factor H related CFHR1 gene deletion is restricted to specific subgroups of patients with atypical haemolytic uraemic syndrome
  1. M-A Dragon-Durey1,2,3,4,
  2. C Blanc2,3,4,
  3. F Marliot1,
  4. C Loirat5,
  5. J Blouin1,
  6. C Sautes-Fridman2,3,4,
  7. W H Fridman1,2,3,4,
  8. V Frémeaux-Bacchi1,2,3,4
  1. 1
    Assistance Publique-Hôpitaux de Paris, Hôpital Européen Georges Pompidou, Service d’Immunologie Biologique, Paris, France
  2. 2
    Université Paris Descartes, Faculté de médecine, Paris, France
  3. 3
    Unité INSERM UMRS 872, Centre de Recherche des Cordeliers, team 13, Paris, France
  4. 4
    Université Pierre et Marie Curie, Paris, France
  5. 5
    Assistance Publique-Hôpitaux de Paris, Hôpital Robert Debré, Service de Néphrologie, Université Paris 7, Paris, France
  1. Dr M-A Dragon-Durey, Service d’Immunologie Biologique, Hôpital Européen Georges Pompidou, 20 rue Leblanc, 75 015 Paris, France; marie-agnes.durey{at}egp.ap-hop-paris.fr

Abstract

Background: Deletion of the complement factor H related 1 (CFHR1) gene is a consequence of non-allelic homologous recombination and has been reported to be more frequent in atypical haemolytic uraemic syndrome (aHUS) patients than in the normal population. Therefore, it is considered a susceptibility factor for the disease. aHUS is associated with hereditary or acquired abnormalities that lead to uncontrolled alternative pathway complement activation. We tested the CFHR1 deletion for association with aHUS in a population of French aHUS cases and controls. Furthermore, we examined the effect of the deletion in the context of known aHUS risk factors.

Methods and results: 177 aHUS patients and 70 healthy donors were studied. The number of CFHR1 alleles was quantified by multiplex ligation dependant probe amplification (MLPA). The frequency of the deleted allele was significantly higher in aHUS patients than in controls (22.7% vs 8.2%, p<0.001). The highest frequency was in the subgroup of patients exhibiting anti-factor H (FH) autoantibodies (92.9%, p<0.0001 vs controls) and in the group of patients exhibiting a factor I (CFI) gene mutation (31.8%, p<0.001 vs controls). The CFHR1 deletion was not significantly more frequent in the cohort of aHUS patients when patients with anti-FH IgG or CFI mutation were excluded.

Conclusions: The high frequency of CFHR1 deletion in aHUS patients is restricted to the subgroups of patients presenting with anti-FH autoantibodies or, to a lesser degree, CFI mutation. These results suggest that the CFHR1 deletion plays a secondary role in susceptibility to aHUS.

View Full Text

Statistics from Altmetric.com

Atypical, or non-shigatoxin associated, haemolytic uraemic syndrome (aHUS, OMIM 235400) is a rare form of thrombotic microangiopathy that principally affects the kidneys’ microvasculature, leading to severe acute renal failure. Genetic susceptibility factors have been identified in almost 60% of aHUS cases. These factors are genetic abnormalities leading to loss-of-function proteins encoded by genes involved in the complement alternative pathway regulation: factor H (CFH), membrane cofactor protein (MCP or CD46) and factor I (CFI).1 Recently, other mutations have been found that lead to gain-of-function proteins encoded by genes for the C3 convertase components, C3 or factor B (CFB).2 3 In addition, a genetic abnormality has been identified that is due to unequal recombination in the RCA (regulators of complement activation) locus, which comprises several genes encoding proteins implicated in regulation of complement activation. CFH and the genes encoding the five complement factor H related proteins (CFHR1-CFHR5) exhibit a high degree of sequence homology. There are two highly homologous sequences with long interspersed nuclear elements (retrotransposons) within this locus, which favour the occurrence of unequal recombinations.4 This mechanism leads to the production of a hybrid protein secondary to the fusion of the CFH gene encoding factor H and CFHR1 gene, leading to a factor H (FH) functional deficiency.5

Finally, aHUS may occur in the context of an autoimmune disease, with the development of anti-FH autoantibodies leading to an acquired FH deficiency.6 7 All these abnormalities lead to a hyperactive alternative pathway C3 convertase.

In addition to these causative factors, several studies have identified additional susceptibility factors that consist of genetic polymorphisms more frequently observed in aHUS patients than in the control population. These include single nucleotide polymorphisms (SNPs) identified in the CFH, MCP or C4BP genes8 9 or a deletion of CFHR1 and CFHR3 genes resulting from an unequal recombination occurring outside the coding sequences of CFH and CFHR1.4 This deletion of CFHR1–CFHR3 has been reported to be associated with aHUS and also to be protective against age related macular degeneration (AMD; OMIM 603075).10 11 Recently, Jozsi et al reported a high frequency of lack of circulating CFHR1 and CFHR3 proteins in a group of patients with the autoimmune form of aHUS,7 which could be due to the same genetic mechanism.

The aim of our study was to determine the frequency of CFHR1 deletion in a French cohort of aHUS patients and to correlate the deleted allele and the deletion homozygosity frequencies with other susceptibility factors. We show that the high frequency of CFHR1–CFHR3 deletion found in aHUS patients is due to a particularly high frequency of the homozygous deletion in the subgroup of patients with anti-FH autoantibody associated aHUS and, to a lesser degree, a CFI mutation.

PARTICIPANTS AND METHODS

Participants

All patients of the French cohort of aHUS were studied between 2002 and 2007. The criteria for the diagnosis of HUS were thrombocytopenia (<150 ×1012/l), renal dysfunction associated with acute anaemia, and fragmented red cells on blood film. None of the patients exhibited biological criteria of a shigatoxin productive bacteria infection. Patients who developed the illness in a context of infectious disease (HIV, Mycoplasma pneumoniae, Borderellia pertussis, varicella virus), haemopathy, solid neoplasia or organ transplantation (except renal graft) were excluded. The cohort consisted of 86 children and 91 adults. Informed consent was obtained from each patient (or parents in the case of children), and the study was approved by the Ethics Committee of the Assistance Publique-Hôpitaux de Paris. The control group comprised 70 French healthy blood donors.

Complement assays

Measurement of CH50 activity in EDTA plasma samples was performed as previously described. Plasma concentrations of the complement components C4, C3 and factor B (FB) antigens were measured by nephelometry (Dade Behring, Paris La Defense, France). Factor H and factor I antigen concentrations were measured by sensitive enzyme linked immunosorbent assay (ELISA) methods and CD46 membrane expression was determined by flux cytometry as previously described.12 Presence of anti-FH IgG was detected by using an ELISA method as previously described.6

Genomic CFH, CFI and MCP DNA sequencing

For genetic analysis, genomic DNA was extracted from peripheral blood cells and amplified by polymerase chain reaction (PCR) using oligonucleotides flanking each exon of the CFH, FI and MCP genes. Primer sequences, length of the PCR amplified fragments, and temperatures of hybridisation used for each reaction and direct DNA sequencing procedure have been previously described.12 Samples were run on the ABI PRISM 3730 Analyser Capillary electrophoresis system (Applied Biosystems, Courtaboeuf, France) and sequence analyses were performed using the Sequencher software.

Multiplex ligation dependant probe amplification

The multiplex ligation dependant probe amplification (MLPA) reaction was performed as previously described.13 Briefly, 50 ng DNA was incubated with 2 fmol of each set of two synthetic probes that hybridise immediately adjacent targets at the sequences of interest. Sequences of probes were designed to determine dosage for exon 5 of CFHR1 and exon 23 of CFH along with control probe C1INH exon 8. In some patients, dosage for exon 3 of CFHR3 was determined. Hybridisation sequences were: CFHR1a: 5′-GACTGACTGAGGACAGCCAAACAGAAGCTTTATTT-3′, CFHR1b: 5′-GAGAACAGGTGAATCAGCTGAATTTG-3′, CFHex23a: 5′-GGACAGCCAAACAGAAGCTTTATTC-3′, CFHex23b: 5′-GAGAACAGGTGAATCAG TTGAATTTG-3′, C1Iex8a: 5′-CTGAAGGGCTTCACGACCAAAGGTGT-3′, C1Iex8b: 5′-CACCTCAGTCTCTCAGATCTTCCCAC-3′, CFHR3a: 5′-GTTTGTACAGGGT AACTCTAC-3′, CFHR3b: 5′-AGAAGTTGCCTGCCATCCTGGC-3′. Probes contained binding sites for primers used for MLPA and a stuffer sequence used to determine a unique length of each amplified probe product. Right hand probes were 5′ phosphorylated (MWG, Roissy, France). MLPA reagents were purchased from MRC Holland (MRC Holland, Amsterdam, The Netherlands), and the reaction was carried out according to the manufacturer’s recommended protocol. Amplified products were diluted 1/10 in deionised formamide (Applied Biosystems) with a ROX 400HD (Applied Biosystems) internal size standard. Samples were run on the ABI PRISM 3730 Analyser Capillary electrophoresis system (Applied Biosystems). Peaks and areas for each sample were determined using Genemapper v4.0 Software (Applied Biosystems) and dosage quotients were calculated.

Statistical analysis

The statistical analysis used the χ2 methods.

RESULTS

We used MLPA reaction to quantify the number of allele copies of CFHR1 gene in a French cohort of 177 patients with aHUS.

A genetic susceptibility factor was identified in 117 (66%) of patients. A genetic abnormality implicating CFH (comprising one CFH/CFHR1 hybrid gene), CFI or MCP was found in 21.5%, 12.4% and 9.6% of patients, respectively. In addition, 16 (9%) patients had a C3 mutation and seven (4%) patients had combined mutations implicating two or three genes. Some of the mutations identified have been reported previously.2 12 1417 Finally, 14 (7.9%) patients presented with an autoimmune form of aHUS with anti-FH autoantibodies.

A genomic deletion of CFHR1 was found in 56 (31.6%) patients. This deletion was homozygous in 12.4% (n = 22) or heterozygous in 19.8% (n = 35) of patients, as determined by the number of CFHR1 copies (table 1).

Table 1 Determination of the number of CFHR1 alleles according the previously identified susceptibility factor in the French atypical haemolytic uraemic syndrome (aHUS) cohort (n = 177 patients)

Thus, the calculated frequency of the deleted allele was 22.7%. In the control population, a genomic deletion was found in 10 controls (14.3%, homozygous in two and heterozygous in eight), giving an allele frequency of 8.2%. Thus, the CFHR1 deleted allele frequency was significantly higher in aHUS patients compared to controls (22.7% vs 8.2%, p<0.001).

The distribution of the deleted allele in our cohort of aHUS patients, according to the susceptibility factor(s) identified in each patient, is depicted in table 2.

Table 2 Frequency of a CFHR1 deleted allele according to the susceptibility factor and comparison of each subgroup with the control population using a χ2 test

The frequency of CFHR1 deletion was particularly high in the subgroup of patients presenting with an autoimmune form of aHUS, in which the allele frequency was 92.9% (χ2 = 94.7, p<0.0001). Among the patients with a genetic abnormality of the complement genes, only those exhibiting a CFI mutation (31.8%, p<0.001) or a C3 mutation (21.9%, p = 0.03) showed a significantly higher frequency of the deleted allele compared to controls. In the other subgroups, no significant difference was observed compared to controls. In patients in whom no genetic factor was identified, the frequency of the deleted allele was significantly higher than in the control population (18.3%, p = 0.02). On the contrary, the frequency of the deleted allele was lower in subgroups of patients exhibiting a genetic abnormality in the CFH and MCP genes than in the control group (7.9% and 2.9%, respectively, vs 8.6% in controls) but did not confer a significant protective status (odds ratio (OR) 0.75, 95% confidence interval (CI) 0.25 to 2.21, p>0.1) (fig 1). After exclusion of patients with anti-FH IgG and CFI mutation, the frequency of the CFHR1 deletion in the cohort of aHUS patients was not significantly different from the frequency in the control group.

Figure 1

Risk for developing atypical haemolytic uraemic syndrome (aHUS) associated with CFHR1 deletion homozygosity. The odds ratio (OR) values are represented by circles. The risk was significant in the subgroups of patients with anti-factor H (FH) autoantibodies and CFI mutation. In two groups of patients, no odds ratio may be calculated (N/A: no available) as no CFHR1 homozygous deletion was found in these groups. CI, confidence interval.

We then hypothesised that CFHR1 deletion homozygosity alone could be a susceptibility factor. Therefore, we studied the distribution of homozygous CFHR1 deletion in the different subgroups of patients according to susceptibility factor(s) (table 3).

Table 3 Frequency of a homozygous CFHR1 deletion according to the susceptibility factor and comparison of each subgroup with the control population using a χ2 test

The frequency of homozygous CFHR1 deletion was higher than in the control population only in the subgroup of patients with CFI mutation (15% vs 2.9% in controls, p = 0.011) and in patients exhibiting anti-FH autoantibodies (92.9% vs 2.9% in controls, p<0.0001). In this last subgroup, 13 of 14 patients exhibited homozygous CFHR1 deletion. The frequency of the deletion homozygosity in the subgroup of patients with no known susceptibility factor was not significantly higher than in controls (5.5% vs 2.9%, OR 1.96, 95% CI 0.32 to 11.8, p>0.1) (fig 1).

We then performed CFHR3 allele quantification by MPLA in the subgroup of patients with anti-FH IgG autoantibodies. With one exception, a complete absence of the CFHR3 gene was observed in all CFHR1 deletion patients.

DISCUSSION

In our study, the deleted allele’s frequency is 8.2% and 22.7% in the normal and the study population, respectively, with deletion homozygosity frequencies of 2.9% and 12.4%. These results are in accordance with the previously reported frequencies in healthy controls and in aHUS patients. The frequency of the CFHR1 homozygous deletion in different control populations has been estimated at 2–5.7%.4 10 18 Zipfel et al reported a frequency of homozygous deletion of CFHR1 in 16% and 10.6% of aHUS patients from the Jena and Newcastle cohorts, respectively.4 In addition, Hughes et al reported the presence of the deleted allele in 20% of chromosomes in a normal but elderly population (n = 170). The frequency was significantly lower (8%) in an age matched population (n = 173) affected by AMD, suggesting that the CFHR1 deletion confers protection against AMD (OR 0.4, 95% CI 0.3 to 0.5).11

The fact that the CFHR1 deletion occurred at a higher frequency in the subgroups of patients exhibiting a mutation in a gene located outside the RCA locus (CFI, C3) suggests that a defect affecting one gene in the RCA locus might be necessary for the development of the disease or play a role in its severity. These roles remain to be elucidated.

The deletion homozygosity is particularly high in the subgroup of patients exhibiting anti-FH autoantibodies. This group comprises 11 children (age at disease onset: 7 months to 13 years; median: 9 years old; five males and six females) and three adults (median age at disease onset: 28 years old; all males). The deletion affects the CFHR1 and CFHR3 genes. These results confirm that the homozygous deletion of CFHR1 and CFHR3 genes is the genetic mechanism responsible for the absence of circulating CFHR1 and CFHR3 observed in the particular group of patients exhibiting an autoimmune form of aHUS.7

The particular correlation that exists between the presence of a homozygous deletion of CFHR1 and CFHR3 genes and the development of anti-FH autoantibodies reveals that most, if not all, patients with an “acquired” form of aHUS share the same homozygous genetic polymorphism. This genetic predisposition seems to be necessary but not sufficient for developing the disease, as this homozygous deletion is also observed in controls. The mechanisms responsible for development of autoantibodies against FH are not yet understood. Our results suggest a particular role of CFHR1 and CFHR3 proteins in the development of this autoimmunity.

In the group of patients without genetic or acquired abnormality affecting the alternative pathway (n = 55), the frequency of the deletion homozygosity is no different than in the control population conferring a non-significant risk for the disease. These results suggest that in our study population, the CFHR1 deletion does not represent a susceptibility factor for aHUS by itself.

In conclusion, the high frequency of CFHR1 deletion observed in our aHUS patients is due to the presence of CFHR1 homozygous deletion in only two categories of patients: those with anti-FH autoantibodies associated HUS and, to a lesser degree, those with a CFI mutation. The results highlight the link between the CFHR1 and CFHR3 homozygous deletion and the development of anti-FH autoantibodies.

REFERENCES

View Abstract

Footnotes

  • Funding: This work was supported in part by grants from the Direction de la Recherche Clinique (DRC) of the Assistance Publique-Hôpitaux de Paris (APHP) (PHRC AOM05130; AOM08198; CIRC 06037) and from GIS-Institut des maladies rares (Maladies Rares et Anomalies de structure du génome).

  • Competing interests: None.

  • Patient consent: Obtained.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.