In Silico Prioritization of Variants of Uncertain Significance in ABCA4 Reveals Conserved Functional Motifs
TL;DR
This integrative computational framework identifies four conserved motifs likely critical to ECD2 and ABCA4 function, and yields a prioritized, mechanistically grounded list of ECD2 variants for targeted experimental validation.
Abstract
Background: ABCA4 variants are the primary cause of Stargardt disease and also contribute to other inherited retinal disorders. Despite this central role, nearly half of all ABCA4 missense variants remain classified as variants of uncertain significance (VUS), limiting genetic diagnosis for many patients. The extracytoplasmic domain 2 (ECD2) harbors a disproportionate share of these unresolved variants yet remains poorly characterized. Methods: We analyzed 244 ECD2 missense variants using a consensus variant effect predictor (VEP) framework integrating PolyPhen-2, SIFT, and MutationTaster. Variants predicted pathogenic by all three tools were designated PAAT (pathogenic across all tools) and prioritized for further evaluation. PAAT variants were assessed using thermodynamic stability modeling ΔΔG and structural analysis, and alignment across ABCA family members was used to identify conserved regions, termed critical conserved motifs (CCMs). Results: PAAT variants were enriched in the central region of ECD2 (Pearson chi-square test, p = 0.0309) and showed a significantly higher frequency of destabilizing ΔΔG values than other variant categories (Cochran-Armitage trend test, p < 0.0001). PAAT variants had 2.2-fold higher odds of occurring within the four identified CCMs (95%: 1.29–3.81), directly linking predicted pathogenicity to sequence conservation. Structural modeling of select variants revealed specific disruptive mechanisms, including steric clashes and altered hydrogen bonding near the retinoid-binding interface. Conclusions: This integrative computational framework identifies four conserved motifs likely critical to ECD2 and ABCA4 function, and yields a prioritized, mechanistically grounded list of ECD2 variants for targeted experimental validation. These findings offer a practical path toward reclassifying ABCA4 VUS and improving genetic diagnosis for patients with Stargardt disease and related inherited retinal disorders.