Identification of Therapeutic Targets for Hyperuricemia: Systematic Genome-Wide Mendelian Randomization and Colocalization Analysis

Abstract

Background: At present, there are still limitations and challenges in the treatment of hyperuricemia (HUA). Mendelian randomization (MR) has been widely used to identify new therapeutic targets. Therefore, we conducted a systematic druggable genome-wide MR to explore potential therapeutic targets and drugs for HUA. Methods: We integrated druggable genome data; blood, kidney, and intestinal expression quantitative trait loci (eQTLs); and HUA-associated genome-wide association study (GWAS) data to analyze the potential causal relationships between drug target genes and HUA using the MR method. Summary-data-based MR (SMR) analysis and Bayesian colocalization were used to assess causality. In addition, we conducted phenome-wide association studies, protein network construction, and enrichment analysis of significant targets to evaluate their biological functions and potential side effects. Finally, we performed drug prediction and molecular docking to identify potential drugs targeting these genes for HUA treatment. Results: Overall, we identified 22 druggable genes significantly associated with HUA through MR, SMR, and colocalization analyses. Among them, two prior druggable genes (ADORA2B and NDUFC2) reached statistically significant levels in at least two tissues in the blood, kidney, and intestine. Further results from phenome-wide studies revealed that there were no potential side effects of ADORA2B or NDUFC2. Moreover, we screened 15 potential drugs targeting the 22 druggable genes that could serve as candidates for HUA drug development. Conclusions: This study provides genetic evidence supporting the potential benefits of targeting 22 druggable genes for HUA treatment, offering new insights into the development of targeted drugs for HUA.

Keywords: Mendelian randomization; colocalization; druggable genes; hyperuricemia; summary-data-based Mendelian randomization.

Figure 1

Overview of the study design [26]. DGIdb: Drug–Gene Interaction Database; eQTLs: expression quantitative trait loci; MR: Mendelian randomization; GWAS: genome-wide association study; SMR: summary-data-based Mendelian randomization; PPI: protein–protein interaction.

Figure 2

Volcano plots for Mendelian randomization (MR) results. (A) Volcano plot for MR results between blood eQTLs and serum uric acid levels. (B) Volcano plot for MR results between kidney eQTLs and serum uric acid levels. (C) Volcano plot for MR results between intestine eQTLs and serum uric acid levels.

Figure 3

Forest plots for six significant Mendelian randomization (MR) result genes. (A) Forest plot for MR results between blood eQTLs and serum uric acid levels. (B) Forest plot for MR results between kidney eQTLs and serum uric acid levels. (C) Forest plot for MR results between intestine eQTLs and serum uric acid levels. The positions of the black squares (coordinates on the horizontal axis) represent the effect size; the horizontal line extending on either side of the black square indicates the confidence interval (usually 95% confidence interval) for the effect value.

Figure 4

Regional genomic plots for conventional colocalization analysis. Colocalizations between serum uric acid levels and (A) blood ADORA2B; (B) blood NDUFC2; (C) kidney ADORA2B; and (D) intestine NDUFC2.

Figure 5

Network pharmacology analysis of the candidate druggable genes. (A) Protein–protein interaction (PPI) network analysis; (B) Gene Ontology (GO) biological function analysis; (C) Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis.

Figure 6

Molecular docking results for the candidate drugs and proteins with binding energies less than −5 kcal/mol.