Investigating Candidate Genes and Disease Mechanism for the Childhood Blinding Disease Familial Exudative Vitreoretinopathy

Abstract

Familial exudative vitreoretinopathy (FEVR) is a childhood blinding disorder that causes incomplete vascularization of the retina. Though the Norrin/FZD4 signalling pathway is established as the disease mechanism of FEVR, the specific interaction of the pathway in retinal angiogenesis is unclear. The aim of this thesis is to expand the FEVR genotype and further characterize the Norrin/FZD4 signalling pathway in the context of retinal vascular development. I determine the pathogenic variant frequency of the FEVR genes FZD4, NDP, LRP5, TSPAN12, and KIF11 in a cohort of FEVR patients. I show that the use of an expanded FEVR gene panel containing all 25 genes with a connection to FEVR does not increase the diagnostic rate for patients without a molecular diagnosis. I establish Norrin/FZD4 binding as the exclusive disease mechanism of Norrie disease. A young girl presenting with a Norrie disease phenocopy was determined to have biallelic FZD4 variants, rendering FZD4 functionally knocked out. The loss of FZD4 resulted in the same phenotype as loss of NDP, underscoring the centrality of Norrin ligand binding to Frizzled-4 receptor in FEVR pathogenesis. Norrin/FZD4 is implicated in retinal angiogenesis, and with human retinal microvascular endothelial cell assays I determine that Norrin contributes to morphogenesis but not cell migration. With Norrin/FZD4 signalling implicated in vessel stabilization and barriergenesis, I establish for the first time the downstream target genes for Norrin/FZD4 signalling at different stages of development. Norrin/FZD4 signalling is bimodal, with the early response targeted towards deactivating Notch signalling. Late-stage Norrin activation targets are involved in formation of junctional complexes to establish the blood retinal barrier. I show the importance of re-evaluating historic sequencing and variant interpretation. The re- analysis of a cohort collected over a two-decade period resulted in the identification of novel variants in FEVR genes. Finally, I demonstrate that FEVR candidate genes should be carefully evaluated with the up-to-date variant classification guidelines. DVL3 was identified as a candidate gene in 2019, but after five years of improving technology and expanding databases, the identified patient-derived DVL3 variants were debunked as potentially FEVR-causal. All together, these findings have expanded upon the role of Norrin/FZD4 signalling in retinal angiogenesis.