Genome-Wide Association Study Reveals Variants in CFH and CFHR4 Associated with Systemic Complement Activation: Implications in Age-Related Macular Degeneration

Abstract

Purpose: To identify genetic variants associated with complement activation, which may help to select age-related macular degeneration (AMD) patients for complement-inhibiting therapies.

Design: Genome-wide association study (GWAS) followed by replication and meta-analysis.

Participants: AMD patients and controls (n = 2245).

Methods: A GWAS on serum C3d-to-C3 ratio was performed in 1548 AMD patients and controls. For replication and meta-analysis, 697 additional individuals were genotyped. A model for complement activation including genetic and non-genetic factors was built, and the variance explained was estimated. Haplotype analysis was performed for 8 SNPs across the CFH/CFHR locus. Association with AMD was performed for the variants and haplotypes found to influence complement activation.

Main outcome measures: Normalized C3d/C3 ratio as a measure of systemic complement activation.

Results: Complement activation was associated independently with rs3753396 located in CFH (P_discovery = 1.09 × 10^-15; P_meta = 3.66 × 10^-21; β = 0.141; standard error [SE] = 0.015) and rs6685931 located in CFHR4 (P_discovery = 8.18 × 10^-7; P_meta = 6.32 × 10^-8; β = 0.054; SE = 0.010). A model including age, AMD disease status, body mass index, triglycerides, rs3753396, rs6685931, and previously identified SNPs explained 18.7% of the variability in complement activation. Haplotype analysis revealed 3 haplotypes (H1-2 and H6 containing rs6685931 and H3 containing rs3753396) associated with complement activation. Haplotypes H3 and H6 conferred stronger effects on complement activation compared with the single variants (P = 2.53 × 10^-14; β = 0.183; SE = 0.024; and P = 4.28 × 10^-4; β = 0.144; SE = 0.041; respectively). Association analyses with AMD revealed that SNP rs6685931 and haplotype H1-2 containing rs6685931 were associated with a risk for AMD development, whereas SNP rs3753396 and haplotypes H3 and H6 were not.

Conclusions: The SNP rs3753396 in CFH and SNP rs6685931 in CFHR4 are associated with systemic complement activation levels. The SNP rs6685931 in CFHR4 and its linked haplotype H1-2 also conferred a risk for AMD development, and therefore could be used to identify AMD patients who would benefit most from complement-inhibiting therapies.