Atypical haemolytic uraemic syndrome associated with a hybrid complement gene

Abstract

Background: Sequence analysis of the regulators of complement activation (RCA) cluster of genes at chromosome position 1q32 shows evidence of several large genomic duplications. These duplications have resulted in a high degree of sequence identity between the gene for factor H (CFH) and the genes for the five factor H-related proteins (CFHL1-5; aliases CFHR1-5). CFH mutations have been described in association with atypical haemolytic uraemic syndrome (aHUS). The majority of the mutations are missense changes that cluster in the C-terminal region and impair the ability of factor H to regulate surface-bound C3b. Some have arisen as a result of gene conversion between CFH and CFHL1. In this study we tested the hypothesis that nonallelic homologous recombination between low-copy repeats in the RCA cluster could result in the formation of a hybrid CFH/CFHL1 gene that predisposes to the development of aHUS.

Methods and findings: In a family with many cases of aHUS that segregate with the RCA cluster we used cDNA analysis, gene sequencing, and Southern blotting to show that affected individuals carry a heterozygous CFH/CFHL1 hybrid gene in which exons 1-21 are derived from CFH and exons 22/23 from CFHL1. This hybrid encodes a protein product identical to a functionally significant CFH mutant (c.3572C>T, S1191L and c.3590T>C, V1197A) that has been previously described in association with aHUS.

Conclusions: CFH mutation screening is recommended in all aHUS patients prior to renal transplantation because of the high risk of disease recurrence post-transplant in those known to have a CFH mutation. Because of our finding it will be necessary to implement additional screening strategies that will detect a hybrid CFH/CFHL1 gene.

Figure 1. Genomic Structure of the Region of the RCA Cluster Containing the Genes Encoding Factor H and the Five factor H-Related Proteins

Genomic duplications including the different exons of the six genes were originally determined by Male et al. [4] and are colour-coded. Exons are indicated as vertical lines. (A) Position of the genes encoding factor H and the factor H-related proteins in a centromeric segment of the RCA cluster at 1q32. (B and C) Nonhomologous recombination occurring at X would result in a hybrid gene consisting of the first 21 exons of CFH (encoding SCRs 1–18 of the hybrid gene) and the last 2 exons of CFHL1 (encoding SCRs 19 and 20 of the hybrid gene). If the recombination occurred at Y this would result in deletion of CFHL3 and CFHL1 but CFH would remain intact. (D) The recombination event would also potentially lead to a duplication of CFHL1 and CFHL3. (Figure adapted from Figure 1 of [15] with kind permission of Human Mutation C 2006, Wiley Liss Inc., A Wiley Company.)

Figure 2. Pedigree of the Family

Figure 3. cDNA Evidence of a Hybrid Gene

(A) Inverted CFH exons 20–23 cDNA sequence showing the site of the first-round forward primers (blue), the second round forward primers (green), and the reverse primers for both rounds (red). The nucleotides at which the CFH and CFHL1 sequences differ are shown in bold and highlighted (excluding exon 20). (B) Inverted cDNA sequence of CFH exons 20–23 from III:6 (affected family member) showing evidence for a hybrid CFH/CFHL1 gene. Positions at which the CFH and CFHL1 sequences differ are indicated by arrows. At the three differences in exons 22 and 23 (numbered 1–3) there is a heterozygous base change, one allele being wild-type CFH and the other the equivalent base from CFHL1. (C) Inverted cDNA sequence showing hybrid CFH/CFHL1 sequence (c.3590T>C, V1197A, and c.3572C>T, S1191L) in III:6 (affected) and III:3 (unaffected carrier) compared to normal CFH sequence in III:1 (unaffected), III:7 (unaffected), and a normal unrelated control.

Figure 4. Identifying and Screening the Breakpoint Region

(A) Sequence of unique PCR product generated with specific CFH (forward) and CFHL1 (reverse) primers from III:6 (affected) demonstrating the hybrid product. Unique CFH positions are indicated with black arrows, and unique CFHL1 nucleotides are indicated with red arrows. (B) The genomic sequence of CFH is shown aligned above CFHL1. Exons 21 and 4 of the two genes respectively are highlighted in grey. The primer-binding sites for the PCR are shown in red. The differences visible in intron 21/4 from sequencing the product are highlighted in the standard base colours A (green), C (blue), G (black), and T (red). The breakpoint is within the region underlined. (C) Sequence of the intron between exons 21 and 22 shows a switch from heterozygosity at CFH/CFHL1 unique bases to a CFHL1 sequence in III:3 (unaffected carrier) and III:6 (affected) compared with III:1 (unaffected), III:7 (unaffected), and a normal unrelated control.

Figure 5. MLPA Binding Sites Used to Identify Deletions of CFH Exons 22 and 23

The hybridisation sequence for the 5′ and 3′ probes are shown by red and blue, respectively. The genomic sequence of CFH is shown aligned above CFHL1.

Figure 6. Southern Blot Evidence of Genomic Rearrangement

(A) A Southern blot using a 1.1 kb probe overlying CFH exon 21 and CFHL1 exon 4 hybridised to HindIII-digested DNA (sites shown as arrows) will result in fragments of 11.2 kb from CFH, 1.4 kb from CFHL1, and 8.6 kb from a CFH/CFHL1 hybrid gene. The site of the 1.1 kb probe is indicated above. (B) Southern blot showing an additional 8.6 kb band (indicated by B) in lanes 2 and 3, which represent III:3 (unaffected carrier) and III:6 (affected) compared to lanes 1, 4, and 5, which represent III:1 (unaffected) and individuals with homozygous deletion of CFHL1. Bands at A and C represent fragments of 11.2 kb from CFH and 1.4 kb from CFHL1, respectively. A size ladder is shown to the right with heavy arrows indicating the expected sizes.

Figure 7. Dosage Evidence of a Hybrid Gene

(A) QF-PCR and (B) MLPA dosage quotients showed an apparent heterozygous “deletion” of CFH exons 22 and 23 in II:8 (unaffected carrier), III:3 (unaffected carrier), III:6 (affected), and IV:1 (affected). There was no evidence of a “deletion” in III:1 and III:7 (both unaffected).