Abstract
Psoriasis affects a significant proportion of the worldwide population and causes an extremely heavy psychological and physical burden. The existing therapeutic schemes have many deficiencies such as limited efficacies and various side effects. Therefore, novel ways of treating psoriasis are urgently needed. A large-scale meta-analysis of psoriasis genome-wide association studies (GWAS) totaling 20 105 cases and 842 975 controls was conducted. Based on the GWAS results, Mendelian randomization (MR) analyses were then performed on three cis-protein quantitative trait loci (pQTL) data in blood. Furthermore, druggability verification and mouse knock-out models were utilized to explore the clinical value of screened proteins. We identified 42 genome-wide significant psoriasis risk variants (P < 5 × 10-8), of which 33 were previously unreported. MR analyses unveiled 19 unique circulating proteins that were associated with psoriasis, among which only AIF1, FCGR3A, NEU1, HSPA1A, TNXB, and ABO were the potential proteins that interacted with psoriasis risk after being analyzed with high evidence of colocalization (PP.H4 > 0.9). In addition, AIF1, FCGR3A, and HSPA1A have been finally determined to be feasible therapeutic targets for psoriasis after being confirmed by druggability verification and specific mouse knock-out models. This large-scale GWAS meta-analysis identified 33 new variants for psoriasis. This study announced that AIF1, FCGR3, and HSPA1A were the unexplored but material variants of psoriasis, thus providing novel and valuable targets for psoriasis treatment and broadening new orientation of drug development for psoriasis.