Evaluation of LOXL1 gene polymorphisms in exfoliation syndrome and exfoliation glaucoma

PURPOSE To evaluate genetic susceptibility of lysyl oxidase-like 1 (LOXL1) gene polymorphisms to exfoliation syndrome (XFS) and exfoliation glaucoma (XFG) in a case-control cohort of American and European patients. METHODS DNA from a total of 620 individuals including 287 exfoliation patients and 333 healthy control subjects were extracted by standard methods. Three single nucleotide polymorphisms (SNPs) of rs1048661 (R141L), rs3825942 (G153D), and rs2165241 were genotyped in these individuals by SNaPshot Assay. The seven coding exons of the LOXL1 gene and their immediate flanking regions were directly sequenced in 95 affected patients. Data management and case-control association studies were performed with SNP-STAT and PLINK programs. The obtained DNA sequences were evaluated with the STADEN package. RESULTS The 287 unrelated exfoliation cases comprised of 171 American patients (mostly of European background) and 116 patients from 12 European countries. This phenotype was further divided into patients with exfoliation only and no glaucoma (XFO; n=95), exfoliation with glaucoma (XFG; n=133), and exfoliation unclassified (XFU; n=59). Genotypic data were analyzed separately for XFO, XFG, XFU, and XFS (all exfoliations; n=287) and for Americans and Europeans. The observed genotypic frequencies for each exfoliation phenotype or population were tabulated separately and tested for deviation from the Hardy-Weinberg equilibrium (HWE) using a standard Chi(2) test. There were no HWE deviations and no significant genotypic differences between these subcategories for the three studied SNPs. For the combined exfoliation cohort, homozygote genotypes of G/G (rs1048661), G/G (rs3825942), and T/T (rs2165241) were significantly overrepresented. Likewise, case-control allelic association for rs1048661 (p=7.74x10(-9)), rs3825942 (p=3.10x10(-17)), and rs2165241 (p=4.85x10(-24)) were highly significant. The corresponding two-locus haplotype frequencies of GG for rs1048661-rs3825942 (p=1.47x10(-27)), GT for rs1048661-rs2165241 (p=1.29x10(-24)), and GT for rs3825942-rs2165241 (p=2.02x10(-24)) were highly associated with exfoliation phenotypes. The combined effect of these three SNPs revealed that the GGT haplotype is overrepresented by 66% in exfoliation cases, and this deviation from controls is highly significant (p=1.93x10(-24)). This haplotype constituted a major risk factor for development of exfoliation in both XFS and XFG. By contrast, the GAC haplotype was significantly underrepresented (p=4.99x10(-18)) in exfoliation cases by 83% and may potentially have a protective effect for this condition with an estimated attributable risk percent reduction of 457%. The only other haplotype that was significantly different between cases and controls was TGC (p=5.82x10(-9)). No observation was made for the GAT haplotype. The combined three haplotypes of GGT, GAC, and TGC were associated with 91% of the exfoliation syndrome cases in the studied populations. Seven coding exons of LOXL1 were also sequenced in 95 affected cases. In addition to the three above-mentioned SNPs, 12 other variations were also observed in these patients (G240G, D292D, A320A, V385V, rs2304719, IVS3+23C>T, IVS3-155G>A, IVS3-101G>A, IVS4+49G>A, rs2304721, IVS5-121C>T, and rs2304722). None were considered a disease-causing mutation. CONCLUSIONS We confirmed a strong association with LOXL1 variants in our patients. For the LOXL1 gene, individual alleles of rs1048661 (G), rs3825942 (G), and rs2165241 (T) are highly associated with XFS and XFG in American and European populations. The GGT haplotype constitutes a major risk haplotype for exfoliation, and GAC may have a protective role. DNA sequencing of 95 affected patients did not show any mutations in this gene. The LOXL1 SNPs are located in the 15q24.1 band and within a genetic locus (GLC1N) that is associated with primary open-angle glaucoma (POAG). However, the LOXL1 genetic predisposition is only limited to exfoliation with or without glaucoma and does not include the POAG phenotype.