Cross-tissue transcriptome-wide association study reveals novel psoriasis susceptibility genes

Abstract

Background

Psoriasis is a chronic, immune-mediated inflammatory skin disorder with a strong genetic component. Although numerous GWAS have identified risk loci, many associated variants lie in non-coding regions, complicating functional interpretation.

Objective

This study aimed to identify novel psoriasis susceptibility genes by integrating large-scale GWAS and eQTL data using a cross-tissue TWAS approach.

Methods

We integrated psoriasis GWAS summary statistics from the FinnGen database with GTEx V8 eQTL data. A cross-tissue TWAS was performed using UTMOST, followed by validation with single-tissue TWAS via FUSION. Conditional and joint analyses were conducted to delineate independent genetic signals, and gene-based analysis was performed using MAGMA. Causal relationships were evaluated using Mendelian randomization (MR) and Bayesian colocalization analyses. Key SNPs were functionally characterized using CADD, GERP++, and RegulomeDB for pathogenicity prediction and regulatory potential assessment. Finally, functional network analysis was conducted using GeneMANIA.

Results

The cross-tissue TWAS identified 259 genes significantly associated with psoriasis (p < 0.05), with 12 remaining significant after FDR correction. Single-tissue TWAS validation revealed 655 significant genes, with an overlap of three protein-coding candidates: POLI, NFKB1, and ZFYVE28. Cross-validation with MAGMA refined the candidate set to NFKB1 and ZFYVE28. MR and colocalization analyses supported a causal relationship for NFKB1 in Skeletal Muscle, Transverse Colon, and Cultured Fibroblasts, and for ZFYVE28 in Subcutaneous Adipose Tissue and Esophageal Mucosa tissues. Functional annotation identified key SNPs including rs4235405, rs3774960, and rs1598856 for NFKB1, and rs1203786 for ZFYVE28, with varying degrees of pathogenicity and regulatory potential. GeneMANIA network analysis further implicated NFKB1 in NF-κB signaling and ZFYVE28 in vesicle-mediated transport.

Conclusion

Our integrative multi-omics approach identifies NFKB1 and ZFYVE28 as novel psoriasis susceptibility genes, providing potential biomarkers and therapeutic targets for this complex disease.