Biomechanical and inflammatory pathways underlying the genetic architecture of keratoconus: A genomic SEM study.

BACKGROUND

The genetic architecture of keratoconus (KC) is complex, and the interplay between its biomechanical and inflammatory etiological components remains poorly defined. We aimed to quantitatively resolve these distinct genetic pathways to provide a mechanistic framework for disease pathogenesis.

METHODS

We applied genomic structural equation modeling (genomic SEM) to large-scale GWAS summary statistics to model two latent genetic factors: a Corneal Biomechanical Stability Factor (CBSF) and an Atopic-Inflammatory Liability (AIL) factor. Pathway-specific multivariate GWAS were subsequently generated, from which candidate causal genes were inferred and then validated for differential expression in two independent patient transcriptomic cohorts. Downstream regulatory effects were interrogated using virtual knockout simulations on single-cell RNA-seq data, followed by pathway enrichment analysis. Functional validation was performed at the protein level using human corneal keratocytes cultured on a soft substrate designed to mimic the KC biomechanical environment.

RESULTS

The genomic SEM supported a dual-pathway architecture, with the CBSF exerting a strong protective effect on KC (β = -0.48, p < 0.01) and the AIL factor conferring significant risk (β = 0.26, p < 0.05), together explaining 32.0 % of KC's genetic variance. From a pool of 112 candidate loci, we prioritized four high-confidence candidate genes. The protective effects of COL1A1, LOXL2, and ITGA1 were traced to their roles in matrix homeostasis, while the risk-conferring activity of SFRP1 was linked to inflammatory signaling. This mechanistic divergence, first revealed through in silico perturbation, was subsequently corroborated at the proteomic level, where our in vitro KC model displayed molecular signatures consistent with the predicted disease state.

CONCLUSION

By resolving the dual genetic pathways of keratoconus, we successfully identified high-confidence candidate genes whose pathogenic relevance was further supported by functional validation. This provides a robust mechanistic framework and identifies high-priority targets for precision therapy.