Metabolomic- Genomic Association in Age-Related Macular Degeneration

Abstract

Purpose : Previous studies have identified genetic variants and metabolites associated with age-related macular degeneration (AMD), a multifactorial disease involving genetic and environmental risk factors. However, the interaction among these factors and how they promote AMD remains largely unknown. Metabolomic-genomic studies can provide unique insights to this end. This study aimed to analyze association between single nucleotide polymorphisms (SNPs) and metabolites in a multicenter cohort of AMD patients and controls.

Methods : Prospectively recruited, cross-sectional, multicenter study (Boston, United States and Coimbra, Portugal). We included 388 subjects with AMD (any stage of disease) and 98 controls without any vitreoretinal disease (age > 50 years). Color fundus photographs were used for AMD grading (AREDS classification scheme). Fasting blood samples were processed and analyzed using ultrahigh-performance liquid chromatography and high-resolution mass spectrometry (Metabolon, Inc.). Single nucleotide polymorphisms (SNP) were genotyped using an Illumina Omni express platform. Association between metabolites and AMD were tested using logistic regression and meta-analysis to combine the two populations. Analyses of metabolomic quantitative trait loci (mQTL) were conducted using linear regression models, adjusted for age, sex, and the 10 first principal components (PC) of both metabolites and SNPs. Benjamini–Hochberg procedures were used to account for false discovery rate (FDR) in all performed analysis.

Results : We first looked at the plasma metabolomic profiles of AMD patients as compared to controls and observed that, accounting for FDR and confounding, 28 metabolites differed significantly between them (FDR<0.05). Among these metabolites, half (n=14) were significantly associated with at least one SNP only in patients with AMD, and not controls (Table 1, p < 1.2e-7). These metabolites corresponded to 890 mQTL (metabolite-SNP pairs), and were mostly amino acids and lipids. N-acetyle asparagine showed the most significant association (690 mQTL).(Figure1)

Conclusions : To our knowledge, this is the first study on metabolomic-genomic associations in AMD. Our results suggest that mQTL SNPs are associated with AMD, thus providing a basis for further exploration. This has the potential to increase our understanding of the biological relevance of AMD-risk SNPs

This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.