TBK1, a prioritized drug repurposing target for amyotrophic lateral sclerosis: evidence from druggable genome Mendelian randomization and pharmacological verification in vitro

Abstract

Background: There is a lack of effective therapeutic strategies for amyotrophic lateral sclerosis (ALS); therefore, drug repurposing might provide a rapid approach to meet the urgent need for treatment.

Methods: To identify therapeutic targets associated with ALS, we conducted Mendelian randomization (MR) analysis and colocalization analysis using cis-eQTL of druggable gene and ALS GWAS data collections to determine annotated druggable gene targets that exhibited significant associations with ALS. By subsequent repurposing drug discovery coupled with inclusion criteria selection, we identified several drug candidates corresponding to their druggable gene targets that have been genetically validated. The pharmacological assays were then conducted to further assess the efficacy of genetics-supported repurposed drugs for potential ALS therapy in various cellular models.

Results: Through MR analysis, we identified potential ALS druggable genes in the blood, including TBK1 [OR 1.30, 95%CI (1.19, 1.42)], TNFSF12 [OR 1.36, 95%CI (1.19, 1.56)], GPX3 [OR 1.28, 95%CI (1.15, 1.43)], TNFSF13 [OR 0.45, 95%CI (0.32, 0.64)], and CD68 [OR 0.38, 95%CI (0.24, 0.58)]. Additionally, we identified potential ALS druggable genes in the brain, including RESP18 [OR 1.11, 95%CI (1.07, 1.16)], GPX3 [OR 0.57, 95%CI (0.48, 0.68)], GDF9 [OR 0.77, 95%CI (0.67, 0.88)], and PTPRN [OR 0.17, 95%CI (0.08, 0.34)]. Among them, TBK1, TNFSF12, RESP18, and GPX3 were confirmed in further colocalization analysis. We identified five drugs with repurposing opportunities targeting TBK1, TNFSF12, and GPX3, namely fostamatinib (R788), amlexanox (AMX), BIIB-023, RG-7212, and glutathione as potential repurposing drugs. R788 and AMX were prioritized due to their genetic supports, safety profiles, and cost-effectiveness evaluation. Further pharmacological analysis revealed that R788 and AMX mitigated neuroinflammation in ALS cell models characterized by overly active cGAS/STING signaling that was induced by MSA-2 or ALS-related toxic proteins (TDP-43 and SOD1), through the inhibition of TBK1 phosphorylation.

Conclusions: Our MR analyses provided genetic evidence supporting TBK1, TNFSF12, RESP18, and GPX3 as druggable genes for ALS treatment. Among the drug candidates targeting the above genes with repurposing opportunities, FDA-approved drug-R788 and AMX served as effective TBK1 inhibitors. The subsequent pharmacological studies validated the potential of R788 and AMX for treating specific ALS subtypes through the inhibition of TBK1 phosphorylation.

Keywords: Amyotrophic lateral sclerosis; Drug repurposing; Druggable gene; Mendelian randomization; TBK1.

Fig. 1

Graphical abstract of genetic instrument variants selection, Mendelian randomization, identification of repurposing drugs discovery, and pharmacological analysis. A Obtaining cis-eQTL data for druggable genes by overlapping discovery cis-eQTL data and confirmatory cis-eQTL data with potentially druggable genes. The SNP data obtained through filtering for cis-eQTL data for blood and brain tissue from eQTLGen and PsychENCODE, with located within ± 100 kb of the TSS and meeting the FDR < 0.05 criteria, is considered as “Discovery cis-eQTL data”. The different sets of SNPs based on various additional selection criteria and multiple eQTL datasets are considered as “Confirmatory cis-eQTL data.” B Workflow showing the MR study and repurposing drugs discovery. Identifying druggable genes with strong genetic associations through the MR and colocalization analysis, further exploring treatment drugs associated with these druggable genes that meet the inclusion criteria. C R788 and AMX inhibit MSA-2-induced cGAS-STING signaling activation. D R788 and AMX inhibit ALS-related toxic protein mediated cGAS/STING signaling. MR Mendelian randomization, cis-eQTL cis-expression quantitative trait loci, ALS amyotrophic lateral sclerosis

Fig. 2

MR and colocalization results of potentially druggable genes. The forest plot displays the results of potential druggable genes under different MR parameters. The bar charts illustrate the colocalization results for each potential druggable gene. Results are color-coded according to tissue (blue = blood, red = brain tissue). FDR false discovery rate, OR odds ratio

Fig. 3

TBK1 was identified as the binding target of FDA-approved drugs. A A summary of the published TBK1 binding affinity measurements of AMX, R406, and R788. B The global (left) and local (right) schematics illustrate the interaction between R406 and SYK by co-crystal structure (PDB: 3FQS) [48]. C Molecular docking model showing the global (left) and local (right) interactions between R406 and TBK1 (PDB: 4IM0) (predicted free energy: − 9.4 kcal/mol) (hydrogen bonds were depicted by solid blue line). D Molecular docking model showing the global (left) and local (right) interactions between R788 and TBK1 (PDB: 4IM0) (predicted free energy: − 8.9 kcal/mol) (hydrogen bonds were depicted by solid blue line, hydrophobic interactions were depicted by black dash lines, halogen bonds were depicted by solid green lines, and salt bridges were depicted by yellow dash lines). E The amino acid sequence comparison between SYK and TBK1 kinase domains

Fig. 4

The cGAS-STING pathway activation induced by MSA-2 is suppressed by R788 and AMX. A The schematics of signaling interactions between R788/AMX and the cGAS-STING pathway. B–C CCK-8 assay showed the cytotoxicities of R788 and AMX with increased dosages. D Western blotting analysis of p-p65, p-IRF3, and p-TBK1 from THP-1 cells subjected to MSA-2 stimulation and treated with R788 (10–20 μM) or AMX (100–200 μM). The qualification of p-TBK1 (E) and p-p65 (F) when normalized with β-actin. G–L RT-qPCR showed the relative levels (normalized with GAPDH) of IFNB, TNFA, and IL6 from THP-1 cells treated with R788 (10 μM) or AMX (200 μM)

Fig. 5

R788 and AMX suppressed cGAS/STING-dependent pro-inflammatory signaling induced by TDP-43/SOD1. A Western blotting analyzed the levels of transiently transfected ALS toxic proteins and p-TBK1 in NSC-34 cells. B WB analysis of p-TBK1, p-p65, and p-IRF3 in doxycycline-inducible TDP43/Q331K NSC-34 cells. C–H RT-qPCR showed the relative levels (normalized with GAPDH) of IFNB, TNFA, and IL6 from NSC-34 cells treated with R788 (10 μM) or AMX (200 μM)