Identification of TMEM106B as a Shared Potential Drug Target for Depression and Stroke Through Comprehensive Genetic...

Background: The coexistence of depression and stroke has long been observed; however, their intrinsic link has not been fully understood. We aimed to inform the importance of depression intervention as a primary prevention of stroke by investigating shared genetic etiology and causal relationship underlying depression and stroke. Methods: Leveraging summary statistics from the hitherto largest genome-wide association studies (GWAS) in European-ancestry individuals for depression (N _case/N _control = 294,322/741,438) and stroke (N _case/N _control = 73,652/1,234,808), we performed cross-trait linkage-disequilibrium (LD) score regression and SUPERGNOVA to quantify global and local genetic correlations, cross-trait meta-analysis to identify shared genetic loci, transcriptome-wide association study (TWAS) to detect shared tissue-specific gene expression, and Mendelian randomization (MR) analysis to make causal inference between the two conditions. Results: We observed a significant positive global genetic correlation between depression and stroke (rg = 0.18, p=2.92 ×10^-9). Partitioning the whole genome, we observed one genomic region (11q23.2) presenting a significant local genetic correlation. Cross-trait meta-analysis and TWAS identified two shared genetic loci (TMEM106B and FES) revealing potential shared biological mechanisms involving lysosome localization. MR identified a putative causal association of genetically predicted depression on stroke (odds ratio [OR] = 1.13, 95% confidence interval [CI] = 1.07-1.19, p=1.12 ×10^-5). A considerable proportion of this association was mediated through smoking initiation (proportion-mediated [PM] = 44.0%, 95% CI = 19.9%-68.1%, p=3.42 ×10^-4), hypertension (PM = 34.0%, 95% CI = 14.5%-53.5%, p=6.46 ×10^-4), type 2 diabetes (PM = 19.0%, 95% CI = 8.5%-29.5%, p=3.78 ×10^-4), and atrial fibrillation (PM = 10.9%, 95% CI = 0.7%-21.1%, p=3.61 ×10^-2), respectively. MR in the reverse direction identified a putative association of genetically predicted stroke on depression (OR = 1.05, 95% CI = 1.01-1.09, p=1.73 ×10^-2), which attenuated to nonsignificant when correcting for both correlated and uncorrelated pleiotropy (OR = 1.00, 95% CI = 0.98-1.03, p=0.88). Drug target MR identified causal associations of genetically predicted TMEM106B level on depression (OR = 0.92, 95% CI = 0.90-0.94, p=2.04 ×10^-12) and stroke (OR = 0.90, 95% CI = 0.86-0.95, p=3.53 ×10^-5). Conclusion: Our work highlights a shared genetic basis and a putative causal relationship between depression and stroke, providing novel insights into the primary prevention of stroke by depression intervention.