OR18-01 Massively Parallel Functional Assay Of Extracellular Insulin Receptor Variants

Abstract

Abstract Disclosure: V. Aslanzadeh: None. G. Brierley: None. G. Kudla: None. R.K. Semple: None. Loss-of-function mutations in the gene for the human insulin receptor, INSR, cause either autosomal dominant or autosomal recessive severe insulin resistance. Biallelic mutations produce extreme insulin resistance which is commonly fatal in infancy or childhood. A wide and growing number of mutations have been associated with disease, with few recurring mutations, and only a small subset have been studied functionally to date. Discriminating loss-of-function pathogenic alleles from insignificant variants is an important barrier to diagnosis and translational studies of novel candidate treatments. Such candidate therapies include a variety of antibody, peptide and small molecule ligands that activate the receptor in a manner distinct to insulin. These show potential to activate mutant receptors that are expressed but show impaired insulin binding or responsiveness. To address this barrier to translation we have now used massively multiplexed, Flow-Seq-based assays simultaneously to determine the effects of 14,000 human INSR extracellular domain missense variants. We assayed receptor cell surface expression, insulin binding and signal transduction using heterologous expression of mutants in mouse embryonic cells in which endogenous Insr and Igf1r genes had been knocked out or knocked down. Preliminary analysis reliably discriminates A. variants that reduce or abolish receptor cell surface expression when measured with two different monoclonal anti-receptor antibodies B. variants with preserved cell surface expression but selectively impaired binding of insulin and C. an overlapping group of variants with severely impaired signalling response to insulin but preserved response to antibody. Our assay confirmed previously studied attributes of a panel of disease-causing variants, and shows strong agreement with prior alanine scanning mutagenesis studies and structural predictions. We thus provide a rich sequence-function map for INSR variant interpretation. A prediction tool trained on known pathogenic INSR alleles is currently being developed for access by the community both to accelerate confident molecular diagnosis of newly identified rare INSR variants, and also to aid stratification of mutations according to potential for activation by non canonical INSR ligands. Presentation: Saturday, June 17, 2023