Revealing Retinal Disease Genes through Genome-wide Screening of Knockout Mice

Abstract

Purpose : Inherited Retinal Diseases (IRDs) affect the retina by causing dysfunction, death, and vision loss. Given their heterogenous nature, the genetic basis of IRDs can be difficult to diagnose in at least a third of patients. This study aims to eluciate the genetics underlying IRDs by identifying candidate IRD genes and determining their role in retinal biology.

Methods : The International Mouse Phenotyping Consortium (IMPC) database of single gene knockout (KO) mice was queried for retinal abnormalities. A comprehensive literature search was performed on all genes associated with abnormal retinal phenotypes to determine known roles in IRDs and retinal biology. Genes without previously published roles in the retina were considered candidate IRD genes. A similar process was used for established IRD genes from the Retinal Information Network (RetNet). Signaling pathways, molecular functions, and protein-protein interaction data of these gene sets were obtained through the Protein Analysis through Evolutionary Relationships (PANTHER) classification system, Kyoto Encyclopedia of Genes and Genomes (KEGG), Reactome, Wikipathways, and String-db.

Results : 572 unique genes were associated with significant (P<0.0001) abnormal retinal phenotypes in IMPC KO mouse lines, 377 of which have no established roles in retinal biology or disease in humans or mice. Validation of abnormal candidate gene phenotypes against those found in independent publications suggests a human disease-mouse model concordance of 87.2%. Pathway analysis of the 377 candidate IRD genes highlighted revealed 43 signaling pathways not yet represented by existing IRD genes. Similarly, we found highly represented pathways such as the mitogen-activated protein kinase (MAPK) signaling pathway. String-db analysis revealed several clusters where an established IRD gene interacts with candidate genes; nine novel clusters composed solely of candidate genes were also found.

Conclusions : A systematic analysis of KO mouse lines identified 377 novel candidate IRD genes that, when knocked out, resulted in a statistically significant abnormal retinal phenotype. Bioinformatic analysis of these candidate genes revealed similar molecular functions between candidate and established IRD genes. They also revealed several pathways and functional clusters unique to candidate genes which may represent novel or underappreciated mechanisms underlying the development of human IRDs.

This abstract was presented at the 2024 ARVO Annual Meeting, held in Seattle, WA, May 5-9, 2024.