Integrating AlphaFold 3 Structural Modeling With Genotype Phenotype Correlations for LRP5 Related Familial Exudative Vitreoretinopathy

Document Type

Conference Proceeding

Publication Date

6-2026

Publication Title

Investigative Ophthalmology and Visual Science

Abstract

Purpose : To combine AlphaFold 3 structural modeling with genotype-phenotype analysis for LRP5 variants in Familial Exudative Vitreoretinopathy (FEVR), evaluating how structural disruption influences severity, age of onset, and treatment burden.

Methods : Retrospective analysis of 38 individuals with LRP5 variants from a prospectively collected cohort, including affected patients and asymptomatic carriers. Variants were classified as splice donor, splice deletion, in-frame deletion, frameshift, or missense. Severity was staged using Pendergast and Trese/IC3D criteria. Visual acuity, surgical interventions, and comorbidities were abstracted. Genotyping used an institutional 8 gene FEVR panel or certified clinical NGS. AlphaFold 3 generated structural models. Predicted disruptions were integrated with NCBI Variant Effect Predictor (VEP) annotations.

Results : Splice donor variants (n=5; NM_002335.3:c.4488+2T>G) caused premature termination removing β catenin signaling areas while preserving the extracellular domain and showed moderate severity (mean Stage 2.4). Splice-region deletions (n=5; c.4488+2_+17del and c.2215+1_+4del) correlated with mild disease (mean Stage 2.25). The c.4488+2T>G allele lies near the transmembrane domain, whereas the c.2215+1_+4del affects β propeller 3. Frameshift variants (n=4) removed the transmembrane domain and intracellular tail, producing the most severe phenotypes (mean Stage 3.63). In frame deletions (n=5) preserved domain architecture and were associated with mild disease courses. Missense variants (n=19), largely within β propellers, ranged from hypomorphic modifiers (Val667Met, Ala1330Val) to pathogenic disruptions (Pro848Gln/Leu, Thr852Met, Arg578Trp), with variable expressivity (mean Stage 2.37). Procedures were required in 47% of patients, highest in frameshift and pathogenic missense groups. Eight carriers were asymptomatic, suggesting incomplete penetrance.

Conclusions : Structural modeling integrated with clinical data clarifies functional classes of LRP5 variants in FEVR: frameshift and splice donor alleles track with early, severe disease; splice/in frame deletions retain partial signaling and milder courses; missense effects depend on structural context. These correlations support risk stratification, counseling, and targeted surveillance.

Volume

67

Issue

7

First Page

OD19

Comments

Association for Research in Vision and Ophthalmology (ARVO) Annual Meeting, May 3-7, 2026, Denver, CO

Last Page

OD19

Share

COinS