Expression and regulation of LOXL1 and elastin-related genes in eyes with exfoliation syndrome (2013)

LOXL1 genetic variants are significantly associated with exfoliation syndrome (XFS), however the impact of the associated variants on disease development is not yet understood. Initially the associated common missense variants, R141L (Arginine 141 Leucine) and G153D (Glycine 153 Asparate), were considered to be pathogenic alleles [1]. ‘Flipping’ of the risk allele in certain populations for both missense variants has provided strong evidence that these missense changes are not biologically significant (Figure 1) [2,3]. Additionally, in an in vitro assay, the associated missense alleles do not affect LOXL1 enzymatic activity [4]. Together, these results suggest that other LOXL1 variant(s), in linkage disequilibrium with the common missense variants, predispose to XFS by affecting protein function and/or gene expression. Several lines of evidence suggest that a reduction in LOXL1 gene expression contributes to disease development and each are discussed in the following sections.

FIGURE 1.

LOXL1 risk alleles in Caucasians, East Asians, and South Africans. LOXL1 exons are shown as solid dark blue rectangles. The 5′ UTR (untranslated region) and 3′ UTR are shown as smaller rectangles. The promoter regulatory region is adjacent to the 5′UTR. The location of the two missense alleles initially associated with exfoliation syndrome R141L (Arginine 141 Leucine) and G153D (Glycine 153 Aspartate) are indicated in exon 1. The allele associated with risk in each of three ethnic groups is shown by larger rectangles directly below exon 1: Caucasians- Arginine at position 141 and Glycine at position 153; East Asians- Leucine at position 141 and Glycine at 153; and South Africans-Arginine at position 141 and Aspartate at position 153. The flipping of risk alleles in different populations indicates that these missense changes do not cause disease but that they are markers of nearby variants (in linkage disequilibrium) that are likely to affect gene expression and/or protein function

Using ocular tissue samples (iris, lens, and ciliary body) from 25 XFS / exfoliation glaucoma (XFG) cases and 25 controls, the expression level and localization of LOXL1, LOXL2, and lysyl oxidase (LOX) were investigated and correlated with the LOXL1 risk genotype and also disease stages (early versus late) [5]. LOXL1 ocular expression was reduced by approximately 20% per risk allele of rs1048661(R141) [5]. An effect on gene expression was not observed for G153D. Additionally, LOXL1 was observed to have higher levels of expression in early disease stages compared to late and this was not correlated with risk genotypes. The expression of LOXL2 and LOX were not altered in XFS- derived tissue samples compared with controls.

The fact that common LOXL1 missense changes (R141L and G153D) are not likely to have biologically relevant effects suggests that other variants in the LOXL1 genomic region must underlie the significant association observed with disease. Haplotype analysis can be used to identify functional variants within a genomic region. Functional variants could include those in coding and/or noncoding regions. To identify variants associated with disease risk, haplotype analysis using variants distributed throughout the LOXL1 regulatory regions and gene coding regions identified a risk haplotype in a Caucasian population from the United States that included a LOXL1 promoter region variant previously shown to reduce gene expression (rs16958477) [6,7]. These variants and other regulatory elements that could influence LOXL1 expression are shown in Figure 2. Of particular interest are the LOXL1-AS1 (antisense RNA) and its promoter, the LOXL1 promoter elements, and a highly conserved region in the distal 5’ LOXL1 regulatory region. Further work will be necessary to determine if these elements impact disease pathogenesis.

FIGURE 2.

LOXL1 5′ regulatory region. Regulatory elements located in the genomic region 5 to the LOXL1 gene are depicted using a modified screen shot from the UCSC genome browzer (http://genome.ucsc.edu/). SNPs of interest include rs12914489 (part of the risk haplotype from reference 6), rs16958477 (promoter region SNP with possible functional consequences, see references 6 and 7), rs1048661 (R141L), and rs3825942 (G153D). The location of the LOXL1-AS1 (LOXL1 antisense 1) transcripts are shown directly above the LOXL1 gene transcript (only LOXL1 exon 1 and part of intron 1 are in this figure). GERP (genomic evolutionary rate profile) scores indicating mammalian conservation for DNA sequence are shown below the gene transcripts. GERP scores >1.0 are considered to be evolutionarily conserved, and conserved sequences are more likely to have a biological impact. Layered histone marks for HSK27Ac for 7 common cell lines are shown below the GERP scores on a scale of 0 to 100. These marks (indicated by peaks) are often found near regulatory elements. DNA methylation sites in two cell lines potentially involved in exfoliation syndrome, iris pigment epithelial cells (HIPEpiC) and nonpigmented ciliary body epithelial cells (NHPCEpiC) are shown in red (highly methylated) and in green (unmethylated). The lack of methylation in the promoter region suggests that the gene is expressed. The LOXL1 promoter region is indicated by the H3K4Me3 marks (shown for 7 common cell types) directly below the methylation marks. DNaseI hypersensitivity sites active in iris pigment epithelial cells and nonpigmented ciliary body epithelial cells are shown in the bottom track. These are regions of open chromatin that could be active regulatory sites.

In addition to LOXL1, several other factors may contribute to XFS pathogenesis, including TGF-β1 (transforming growth factor beta 1), oxidative stress, UV light and hypoxia. Factors potentially influencing XFS pathogenesis were tested for their effect on LOXL1 expression using cultured fibroblasts from Tenon's capsule biopsies from five patients with cataract and XFS and three patients with cataract without XFS [8]. Changes in LOXL1 expression were also investigated for high and low LOXL1 risk haplotypes (based on the R141L and G153D polymorphisms) [8]. Each of these disease-associated factors, (TGF-β1, oxidative stress, UV light and hypoxia), induced significant increases in expression of LOXL1 and elastic proteins in cells with the low-risk haplotype but had comparatively lower expression in cells with the high-risk haplotype. These results suggest that the high-risk LOXL1 haplotype is associated with reduced gene expression at baseline and also under conditions where LOXL1 gene expression increases in response to factors impacting disease pathogenesis.

To investigate the impact of loss of LOXL1 function on disease development, the phenotypic features of a LOXL1 null mouse were characterized [9]. The LOXL1 null mouse was found to have some features of XFS, including lens abnormalities consistent with cataract and disruption of the blood-aqueous barrier [10]. The LOXL1 null mouse did not have deposition of exfoliation fibrillar material or elevated intraocular pressure and glaucoma. These results suggest that loss of LOXL1 enzyme activity contributes to disease predisposition, but that other factors, which may be genetic and/or environmental, are necessary for the disease to be fully manifest.

Collectively, these results support the hypothesis that dysregulation of LOXL1 expression is a contributing factor to exfoliation disease development. Further study defining the regulatory elements necessary for regulation of LOXL1 expression could point to possible therapeutic targets.