Integration of ‘omics data dissects genetic and metabolic drivers of Macular Telangiectasia Type II

Abstract

Macular telangiectasia type 2 (MacTel), is a rare degenerative eye disease. In this thesis, we will explore and analyse MacTel ‘omics data including genomics, metabolomics and phenomics data to discover MacTel causal drivers that might be targeted for therapeutic interventions. The first chapter introduces the disease, the concept of data integration, and different types of ‘omics data. The second chapter presents the first MacTel GWA study which discovered five loci. Four of them indicated a possible involvement of the glycine/serine metabolic pathway on the disease while the last pointed to genetic influences from retinal vasculature calibre. In the same study, glycine, serine and threonine were externally validated as metabolic biomarkers for MacTel. The third chapter presents a causal model for MacTel via genetic variants. Using Mendelian Randomization, we tested the genetic contribution to the disease of more than 140 metabolites and comorbid traits and found causal effects from serine and glycine, whilst eliminating other spurious metabolites like threonine. Serine was also found to cause disease progression based on longitudinal retinal image data for 455 patients. Novel genetic loci were further revealed via conditional GWAS and MTAG. Analyses of serum metabolomics data presented in the fourth chapter revealed widespread metabolic rewiring, reflecting glycine and serine commensurate with genetic disruptions and identified lipids dysregulation as a new risk factor for MacTel. In the fifth chapter, we confirm that genetically induced depletion of serine in MacTel patients likely results in accumulation of deoxy-sphingolipids. These appear to be extremely toxic for retinal and neuronal tissues and were associated with disease progression. Lastly, the sixth chapter summarises and discusses the main findings of these four studies while placing their discoveries on the bigger spectrum of macular and eye disorders as well as suggesting future studies. The results presented in this thesis significantly advance the understanding of MacTel aetiology and are supporting sensitive diagnostic re-evaluation and targeted laboratory experimentation as well as inclusion criteria for upcoming clinical trials.