Abstract: Objective: To investigate novel therapeutic targets for atopic dermatitis (AD). Methods: Mendelian randomization analysis, colocalization analysis, and phenome-wide association study (PheWAS) were conducted using protein quantitative trait locus (pQTL) data of 2,923 proteins from the UK Biobank Pharma Proteomics Project and genome-wide association study (GWAS) data of AD from Finnish databases. Results: Two-sample Mendelian randomization results showed that the expression levels of proteins including HSPB6, LAT, LDLRAP1, FGR, CSF2, INPP5D, MDM1, FKBP1B, MMP12, and IL7R may increase the risk of AD, while the expression levels of IL36A, SF3B4, SEPTIN8, APOA1, BOLA1, SV2A, and B4GAT1 may reduce the risk of AD. Sensitivity analysis was performed for transcriptome-wide Mendelian randomization (TSMR), and summary data-based Mendelian randomization (SMR) analysis indicated that the associations between the plasma expression levels of SV2A, APOA1, SF3B4, BOLA1, INPP5D, MMP12, FGR, HSPB6, LAT, FKBP1B, IL36A and AD were associated with AD independently. Subsequently, Bayesian colocalization analysis revealed that INPP5D, MMP12, LAT, and IL36A shared the same causal variants with AD. Finally, PheWAS demonstrated that these four potential drug targets did not show significant correlations with other traits at the genetic level. Conclusion: INPP5D, MMP12, LAT, and IL36A proteins are associated with the incidence of AD. Drugs targeting these proteins may have fewer adverse reactions, and this study provides a theoretical basis for the development of new therapeutic strategies for AD.

Key words: atopic dermatitis, Mendelian randomization, colocalization analysis, proteome-wide, genome-wide association study