Increased retinal mtDNA damage in the CFH variant associated with age-related macular degeneration

Abstract

Age-related macular degeneration (AMD) is a major cause of blindness among the elderly in the developed world. Genetic analysis of AMD has identified 34 high-risk loci associated with AMD. The genes at these high risk loci belong to diverse biological pathways, suggesting different mechanisms leading to AMD pathogenesis. Thus, therapies targeting a single pathway for all AMD patients will likely not be universally effective. Recent evidence suggests defects in mitochondria (mt) of the retinal pigment epithelium (RPE) may constitute a key pathogenic event in some AMD patients. The purpose of this study is to determine if individuals with a specific genetic background have a greater propensity for mtDNA damage. We used human eyebank tissues from 76 donors with AMD and 42 age-matched controls to determine the extent of mtDNA damage in the RPE that was harvested from the macula using a long extension polymerase chain reaction assay. Genotype analyses were performed for ten common AMD-associated nuclear risk alleles (ARMS2, TNFRSF10A, CFH, C2, C3, APOE, CETP, LIPC, VEGF and COL10A1) and mtDNA haplogroups. Sufficient samples were available for genotype association with mtDNA damage for TNFRSF10A, CFH, CETP, VEGFA, and COL10A1. Our results show that AMD donors carrying the high risk allele for CFH (C) had significantly more mtDNA damage compared with donors having the wild-type genetic profile. The data from an additional 39 donors (12 controls and 27 AMD) genotyped for CFH alleles further supported these findings. Taken together, these studies provide the rationale for a more personalized approach for treating AMD by uncovering a significant correlation between the CFH high risk allele and accelerated mtDNA damage. Patients harboring this genetic risk factor may benefit from therapies that stabilize and protect the mt in the RPE.

Keywords: Age-related macular degeneration; Complement factor H; Eyebank tissue; Haplogroups; Inflammation; Mitochondria; mtDNA.

Figure 1. RPE mtDNA damage increases with AMD progression

Long-extension PCR was used to determine mtDNA lesion frequency in RPE from human donor macula. Data were normalized to the mean of the non-disease age-matched controls (MGS1). Sample size for each group is provided on the graph. (A) Summary of donors from previous studies (Karunadharma et al., 2010; Terluk et al., 2015) used in the current analysis. (B) Graph of mtDNA damage for donors without AMD (MGS1) and donors with progressively more severe stages of AMD using donors from the combined data set. MGS groups were compared by one-way ANOVA and Tukey’s post hoc test. Significance was set at p<0.05. *P<0.001; MGS1 was less than MGS2, MGS3 and MGS4. Data are mean + SEM.

Figure 2. Distribution of control and AMD donors for each genotype

Donors were genotyped for 10 risk variants associated with (A) cell survival (ARMS2, TNFRSF10A), the complement pathway (CFH, C2, C3), (B) lipid transport/metabolism (APOE, CETP, LIPC), angiogenesis (VEGFA) and remodeling of the extracellular matrix (COL10A1). Shaded area of each bar shows the percent of control (MGS1, light grey) and AMD (MGS2, hatched, MGS3, medium grey, MGS4, dark grey) donors that are homozygous non-risk, heterozygous risk, or homozygous risk (underlined) for each gene. The total number of donors in each risk category is provided at the base of each bar.

Figure 3. RPE mtDNA damage and AMD-associated risk variants

Donors were separated into genotypes for five risk variants that are associated with cell survival (TNFRSF10A), the complement pathway (CFH), lipid transport/metabolism (CETP), angiogenesis (VEGFA) and remodeling of the extracellular matrix (COL10A1). Graphs show the amount of mtDNA damage (relative to the average damage of MGS1 control donors) for donors within each genotype. Genotypes are shown above the gene name; the homozygous risk alleles (dark grey bars) are underlined. The number of donors in each group is shown within the bars. mtDNA damage was compared between non-risk, heterozygous risk, and homozygous risk for each gene by one-way ANOVA and Tukey’s post-hoc test when required. The probability values are provided in the graph for each comparison. Significance was set at p=0.04 following the Benjamini-Hochberg correction for multiple comparisons. * Tukey’s post-hoc test showed CC is significantly higher than TT for CFH.

Figure 4. Association Between mtDNA Damage and the CFH Risk Allele. (A)

Refined analysis of data for CFH from Figure 3 shows the distribution of AMD donors (MGS2, 3, 4) for each CFH genotype (left panel) and the mtDNA damage measured in control (n=42) and AMD donors (n=76) (middle panel). A significant linear increase in mtDNA damage correlated with content of the C risk allele in AMD donors (p=0.049) but not in control donors (p=0.59). AMD donors harboring 1 or 2 high risk alleles had significantly higher mtDNA damage compared with AMD donors homozygous for the low risk alleles (TT) (p=0.043; right panel). (B) Distribution of AMD donors (MGS 3 and 4 only) for each genotype (left panel) includes 43 donors from previous studies (Karunadharma et al., 2010; Terluk et al., 2015) and 27 new donors. The extent of mtDNA damage for AMD and control donors for each genotype is shown in the middle panel. While there was a 20% increase in mtDNA damage in donors carrying the C allele, a linear relationship was not observed (p=0.11). Regression analysis of mtDNA damage for control donors (including 12 new donors) was not significant (p=0.09). T-test analysis (right panel) showed AMD donors harboring 1 or 2 high risk alleles had significantly higher mtDNA damage compared with AMD donors homozygous for the low risk alleles (TT) (p=0.022). Total number of donors in each group is indicated on the graph. Data shown are mean ± SEM.

Figure 5. No Correlation Between mtDNA Haplogroups and mtDNA Damage

The number of donors for each comparison is provided in the bars. The p value for the t-test comparison is shown at the top of each graph. (A) Comparison of mtDNA damage for donors with H (light grey) versus JTU haplogroup (dark grey) cluster was not significantly different. (B) No additive risk for mtDNA damage was found when comparing the CFH alleles on the H and JTU cluster backgrounds.