A Variant of the HTRA1 Gene Increases Susceptibility to Age-Related Macular Degeneration

See allHide authors and affiliations

Science  10 Nov 2006:
Vol. 314, Issue 5801, pp. 992-993
DOI: 10.1126/science.1133811


Age-related macular degeneration (AMD) is the most common cause of irreversible vision loss in the developed world and has a strong genetic predisposition. A locus at human chromosome 10q26 affects the risk of AMD, but the precise gene(s) have not been identified. We genotyped 581 AMD cases and 309 normal controls in a Caucasian cohort in Utah. We demonstrate that a single-nucleotide polymorphism, rs11200638, in the promoter region of HTRA1 is the most likely causal variant for AMD at 10q26 and is estimated to confer a population attributable risk of 49.3%. The HTRA1 gene encodes a secreted serine protease. Preliminary analysis of lymphocytes and retinal pigment epithelium from four AMD patients revealed that the risk allele was associated with elevated expression levels of HTRA1 mRNA and protein. We also found that drusen in the eyes of AMD patients were strongly immunolabeled with HTRA1 antibody. Together, these findings support a key role for HTRA1 in AMD susceptibility and identify a potential new pathway for AMD pathogenesis.

Age-related macular degeneration (AMD) is the leading cause of irreversible blindness in developed countries, affecting 10 million people worldwide (1, 2). Early AMD is characterized by the presence of soft drusen in the macula without vision loss. Advanced AMD is associated with vision loss due to either geographic atrophy of retinal pigment epithelium (RPE) and photoreceptors (GA or dry AMD) or neovascular choriocapillary invasion across Bruch's membrane into RPE and photoreceptor layers (wet AMD). The prevalence of AMD is rising as a result of increasing life expectancy, but the etiology remains poorly understood. Both genetic predisposition (3) and environmental factors such as smoking (4) play an important role in AMD pathogenesis.

Previous work has shown that allelic variants of genes encoding members of the alternative complement pathway, including complement factor H (CFH) (59) and factor B/C2, affect an individual's risk of developing AMD (10). In particular, variants of the CFH gene on chromosome 1q31 confer a major risk for AMD (59, 11, 12). Several independent association studies have implicated a second major locus for AMD at chromosome 10q26 (1216). To identify the critical gene at this locus, we genotyped 442 AMD cases and 309 controls in a Caucasian cohort in Utah, using a panel of 15 single-nucleotide polymorphisms (SNPs) centered around the highest risk associated SNP, rs10490924. In agreement with previous reports, rs10490924 was found to have a significant association signal [P =8.1 × 10–8 for an additive allele-dosage model, ORhet = 1.35 (0.99, 1.86), ORhom = 6.09 (3.27, 11.34), T allele: 39.7% in cases versus 24.7% in controls]. However, of the 15 SNPs analyzed, rs11200638 was the most significantly associated variant [P = 1 × 10–9, ORhet = 1.86 (1.35, 2.56), ORhom = 6.56 (3.23, 13.31), A allele: 40.3% in cases versus 25.2% in controls] (Fig. 1A and table S2). In terms of the significance of the association, the TA haplotype across rs10490924 and rs11200638 was superior to rs10490924 (P = 2.2 × 10–9), but inferior to rs11200638.

Fig. 1.

Genetic association (A) and preliminary analysis of HTRA1 mRNA and protein expression in AMD (B and C). (A) Graphed log P-values (y axis) from association analyses for the 15 SNPs at the 10q AMD region using 442 AMD cases and 309 controls (see also table S2). For the rs10490924 (green square) and rs11200638 (red triangle) SNPs, associations were derived from a larger sample size (581 AMD cases). (B) Real-time RT-PCR semiquantitative analysis of HTRA1 RNA levels in blood lymphocytes from three AMD patients with the AA genotype and three normal controls with the GG genotype. The statistical significance of the differences in expression level was examined using an independent samples t test (SPSS version 13.0): AA:GG (P =0.02). Each RT-PCR reaction was run twice, and the error bars indicate the 95.0% confidence interval of the mean. (C) Drusen is immunolabeled with HTRA1 antibody in an eye from a patient with wet AMD. The staining (arrows) occurs between RPE and Bruch's membrane. Similar results were obtained in another two AMD eyes (fig. S1, B and C). Scale bar = 20 μm.

To investigate these associations further, we genotyped an additional 139 AMD patients for these two variants. The results for both SNPs increased in significance (rs10490924, P = 1.2 × 10–8; rs11200638, P =1.6 × 10–11), with variant rs11200638 remaining the best single variant explaining the association [ORhet = 1.90 (1.40, 2.58), ORhom = 7.51 (3.75, 15.04)]. We next considered association analyses based on genotypes at both rs11200638 and the CFH rs1061170 (Y402H) variant at chromosome 1q31. In a global two-locus analysis enumerating all nine two-locus genotype combinations, the association with AMD was significant (χ2 = 56.56, 8 df; P = 2.2 × 10–9). Table 1 shows the risk estimates for each two-locus genotype combination compared with the baseline of no risk genotypes (TT at CFHY402H and GG at rs11200638). The association of rs11200638 to AMD was significant when analyzed conditional on the presence of the CFH C risk allele (P = 5.9 × 10–8). In particular, this conditional analysis indicates an allele-dosage effect such that homozygotes for the A risk allele of rs11200638 are at an increased risk [ORhom = 7.29 (3.18, 16.74)] over that of heterozygotes [ORhet = 1.83 (1.25, 2.68)] in all AMD cases, even when compared with a baseline that includes individuals who carry the risk genotypes at CFH. With an allele-dosage model, the estimated population attributable risk (PAR) for rs11200638 is 49.3%. Consistent with an additive effect, the estimated PAR from a joint model with CFH Y402H (that is, for a risk allele at either locus) is 71.4%.

Table 1.

Two-locus odds ratios for HTRA1 rs11200638 and CFH rs1061170. Odds ratios with 95% confidence intervals in parentheses were calculated to compare each genotypic combination to the baseline of homozygosity for the common allele at both loci (TT/GG).

SNPHTRA1 rs11200638
CFH rs1061170 (Y402H) GG AG AA
TT 1.00 1.80 (0.93,3.49) 3.43 (0.62,19.00)
CT 1.07 (0.59,1.94) 2.31 (1.28,4.17) 7.31 (2.68,19.93)
CC 3.07 (1.50,6.27) 3.97 (1.93,8.15) 31.52 (4.01,247.96)

The SNP rs11200638 is located 512 base pairs (bp) upstream of the transcription start site of the HTRA1 gene (also known as PRSS11, NM_002775). Using MatInspector (Genomatix Software, GmbH, München, Germany), to scan putative transcription factor binding sites within this region, we identified a conserved AP2/SRF binding element that is altered by the A risk allele. To investigate the functional significance of the SNP, we used real-time reverse transcription polymerase chain reaction (RT-PCR) to study the expression levels of HTRA1 mRNA in lymphocytes of four AMD patients carrying the risk allele AA and three normal controls carrying the normal allele GG (Fig. 1B and fig. S1A). The HTRA1 mRNA levels in lymphocytes from AMD patients with the AA genotype were higher by a factor of ∼2.7 than those in normal controls with the GG genotype (Fig. 1B). The mean HTRA1 protein level in RPE of four AMD donor eyes with a homozygous AA risk allele was higher by a factor of 1.7 than that of six normal controls with a homozygous GG allele (fig. S2). Unfortunately, our analysis of human eye tissue was limited because we were able to obtain only four AMD donor eyes with an AA genotype out of the 60 donors for this study; therefore, these data thus far show a trend toward higher expression with the risk AA allele. Immunohistochemistry experiments revealed that HTRA1 immunolabeling is present in the drusen of three AMD patients (Fig. 1C and fig. S1, B and C).

The HTRA1 gene encodes a member of a family of serine proteases expressed in the mouse retina and RPE (17). HTRA1 appears to regulate the degradation of extracellular matrix proteoglycans. This activity is thought to facilitate access of other degradative matrix enzymes, such as collagenases and matrix metalloproteinases, to their substrates (18). Conceivably, overexpression of HTRA1 may alter the integrity of Bruch's membrane, favoring the invasion of choroid capillaries across the extracellular matrix, as occurs in wet AMD. HTRA1 also binds and inhibits transforming growth factor-β (TGF-β), an important regulator of extracellular matrix deposition and angiogenesis (17). DeWan et al. (19) report that the same HTRA1 SNP is associated with a wet AMD phenotype in a Chinese population. Together, these findings support a key role for HTRA1 in AMD susceptibility and identify a potential new pathway for AMD pathogenesis.

Supporting Online Material

Materials and Methods

Figs. S1 and S2

Tables S1 and S2


References and Notes

View Abstract

Navigate This Article