Prioritizing drug targets in systemic lupus erythematosus from a genetic perspective: a druggable genome-wide Mendelian randomization study

Abstract

Objectives: Systemic lupus erythematosus (SLE) is a heterogeneous autoimmune disease with an unsatisfactory state of treatment. We aim to explore novel targets for SLE from a genetic standpoint.

Methods: Cis-expression quantitative trait loci (eQTLs) for whole blood from 31,684 samples provided by the eQTLGen Consortium as well as two large SLE cohorts were utilized for screening and validating genes causally associated with SLE. Colocalization analysis was employed to further investigate whether changes in the expression of risk genes, as indicated by GWAS signals, influence the occurrence and development of SLE. Targets identified for drug development were evaluated for potential side effects using a phenome-wide association study (PheWAS). Based on the multiple databases, we explored the interactions between drugs and genes for drug prediction and the assessment of current medications.

Results: The analysis comprised 5427 druggable genes in total. The two-sample Mendelian randomization (MR) in the discovery phase identified 20 genes causally associated with SLE and validated 8 genes in the replication phase. Colocalization analysis ultimately identified five genes (BLK, HIST1H3H, HSPA1A, IL12A, NEU1) with PPH4 > 0.8. PheWAS further indicated that drugs acting on BLK and IL12A are less likely to have potential side effects, while HSPA1A and NEU1 were associated with other traits. Four genes (BLK, HSPA1A, IL12A, NEU1) have been targeted for drug development in autoimmune diseases and other conditions.

Conclusions: .This study identified five genes as therapeutic targets for SLE. Repurposing and developing drugs targeting these genes is anticipated to improve the existing treatment state for SLE. Key Points • We identified five gene targets of priority for the treatment of SLE, with BLK and IL12A indicating fewer side effects. • Among the existing drugs that target these candidate genes, Ustekinumab, Ebdarokimab, and Briakinumab (targeting the IL12 gene) and CD24FC (targeting HSPA1A) may potentially be repurposed for the treatment of SLE.

Keywords: Drug target; Expression quantitative trait loci; Genetics; Mendelian randomization; Systemic lupus erythematosus.

Fig. 1

The overview of the study design. First, we acquired druggable genes. Second, GO analysis was conducted. Third, we performed two-sample MR in the discovery and replication phases using blood cis-eQTLs and two large SLE cohorts, respectively. Then, colocalization analyses were then conducted to clarify the robustness of the results. Moreover, we assessed the side effects of drugs that target the confirmed genes through PheWAS. Finally, we made drug predictions of targets and analyzed the pharmacogenomic interactions of existing drugs via multiple databases. GO, gene ontology; MR, Mendelian randomization; eQTL, expression quantitative trait loci; SLE, systemic lupus erythematosus; PheWAS, phenome-wide association analysis

Fig. 2

Enrichment bubble map for GO analysis of genes and Manhattan plot for MR analysis in the discovery phase. a Bubble map of 15 biological processes (BP) obtained by GO analysis of candidate genes. Abscissa represents the gene ratio and the ordinate represents the go entry name. Bubble size represents the count of genes and the color represents the corrected P value. b MR results in the discovery phase revealed that 20 genes are significantly causally associated with SLE (FDR < 0.05). These genes are labeled and highlighted with blue dots in the plot. MR, Mendelian randomization; SLE, systemic lupus erythematosus; FDR: false discovery rate

Fig. 3

Forest plots for MR results of genes causally associated with SLE in the discovery and replication phases. a Forest plot for MR results of 20 significant genes identified in the discovery phase. b Forest plot for MR results of 14 significant genes identified in the replication phase. The threshold of significance was set at FDR < 0.05. The red line represents the estimates using the IVW method, and the blue line represents the estimates using the Wald ratio method. MR, Mendelian randomization; SLE, systemic lupus erythematosus; FDR: false discovery rate

Fig. 4

Manhattan plot for PheWAS of blood BLK, HSPA1A, IL12A, and NEU1. Each dot represents a disease trait within a specific category on the horizontal axis, with different colors representing three different genes. The red dashed line represents the significance threshold for the p value at 5e-08. Several phenotypes discussed in detail in this study are labeled