Abstract

Large copy number variants (CNVs) such as deletions in the regulator of complement activation (RCA) gene cluster (CFH, CFHR3, CFHR1, CFHR4, CFHR2, CFHR5) are associated with elevated risk for atypical hemolytic uremic syndrome (aHUS), a complement regulation disorder (Zipfel 2007). Of particular interest are compound heterozygous CFHR3-CFHR1 and CFHR1-CFHR4 deletions, linked to aHUS due to the resulting composite homozygous CFHR1 deletions (Moore 2010). Detection of CFHR4 CNVs is important as it enables identification of said CFHR3-CFHR1 deletion in trans with the CFHR1-CFHR4 deletion, as well as rare reported CFHR4 duplications (Sudmant 2010). Large heterozygous deletions or duplications involving CFHR3 and CFHR1 or CFHR1 and CFHR4 are undetectable by Sanger sequencing and difficult to detect with amplicon-based Next-Generation Sequencing (NGS). When CNVs are detected by NGS, confirmation with an independent reference method is required by regulatory agencies including the New York State Department of Health and College of American Pathology.

Multiplex Ligation-Dependent Probe Amplification (MLPA®) has been used to reliably detect large CNVs for CFH, CFHR3, CFHR1, CFHR2 and CFHR5, and a commercial research use only (RUO) probe mix is available for this purpose from MRC-Holland (Amsterdam, NL).

For the CFHR4 region, however, MLPA® probes are not available from any commercial source. Furthermore, CFHR4 MLPA® probe design is complicated by high internal and external homology with adjacent RCA regions (including 91.3-96.2% homology within CFHR4 exons, as well as 90-100% homology with CFHR3 exons). Using a combination of MLPA® probe design freeware, commercial software, literature sources and in-house modifications, a multiplex MLPA® laboratory-developed test (LDT) targeting 5 regions of interest (ROIs) in the CFHR4 gene was designed and validated.

The multiplex CFHR4 MLPA® LDT was customized to include an internally-designed ROI MLPA® probe mix targeting CFHR4 intron 1, exon 2, exon 5, intron 6, intron 9, as well as commercially available MLPA® quality control and reference probes. Exon 2 probe sequences were obtained from the literature (Kubista 2011). The remaining MLPA® ROI probes were designed using AlleleID® software (Premier Biosoft) or online freeware MAPD-MLPA (courtesy Stony Brook University Bioinformatics). The MLPA®CFHR4 ROI probes were optimized and challenged to achieve complete specificity for and coverage across the CFHR4 region, as well as to prevent overlap with quality control and reference probes.

The CFHR4 MLPA® LDT was validated with 10 genomic DNA samples, comprising a mixture of healthy donors or patients (BloodCenter of Wisconsin archives) and 6 samples from Coriell Repositories (Camden, NJ), all of which were known to harbor large RCA CNVs. 100% concordance with expected findings was observed.

Full aHUS genetic evaluation comprising NGS analysis of 15 genes and MLPA® analysis of the RCA gene cluster has now been performed for 161 patients. In 95 (59%) of these patients, no large CNV was detected. Forty patients (25%) tested negative for any large CFHR4 variant but did test positive for the CFHR3-CFHR1 heterozygous deletion. 28% of the population are reported heterozygous for this CFHR3-CFHR1 deletion (Moore 2010). Fifteen patients (9%) were positive for homozygous CFHR1 deletions, which are associated with increased risk for development of autoantibody to complement factor H (CFH), as reported by Moore 2010. Using the CFHR4 MLPA® LDT, the latter homozygous CFHR1 deletions were attributed to compound heterozygosity for the CFHR3-CFHR1 and CFHR1-CFHR4 deletions in 3 of the aforesaid 15 patients. The other 12 patients were found to harbor homozygous deletions of CFHR3 and CFHR1. In 3 (1.9%) patients, the CFHR4 MLPA® LDT revealed a CFHR1-CFHR4 heterozygous deletion, similar to the previously reported frequency of 1.4% in aHUS patients (Moore 2010). Finally, 2 patients tested positive for a heterozygous CFHR3 deletion and a CFHR4 duplication, consistent with the presence of a CFHR3-CFHR1 deletion on one allele and CFHR1-CFHR4 duplication on the other allele resulting in a normal CFHR1 copy number.

Overall, these data show the importance of CFHR4 MLPA® CNV testing to aid genetic evaluation of suspected aHUS cases, and this testing should be part of the aHUS genetic panel.

Disclosures

Friedman:Alexion: Speakers Bureau; NovoNordisk: Consultancy; Shire: Consultancy; CSL Behring: Consultancy.

Author notes

*

Asterisk with author names denotes non-ASH members.