To identify potential druggable targets for Alzheimer's disease (AD) by analyzing circulating inflammatory proteins using Mendelian randomization (MR). Two-sample MR analysis was employed to investigate the causal relationships between 91 circulating inflammatory proteins and AD. The primary MR method utilized was the inverse variance weighted (IVW) model, while the weighted median (WM) and MR-Egger models were applied for sensitivity analysis. To assess the heterogeneity of instrumental variables (IVs), Cochran's Q-test and I2 statistics were utilized. Additionally, ChEMBL and DGIdb databases with Bayesian colocalization analysis were consulted to identify potential druggable proteins. MR analysis identified eight inflammatory proteins significantly associated with AD risk. Among these proteins, TNFB odds ratio (OR): 1.06, 95% Confidence Interval (CI): 1.02-1.11, p = 8.77×10- 3, TSLP (OR: 1.10, 95% CI: 1.01-1.19, p = 0.028), S100A12 (OR: 1.09, 95% CI: 1.01-1.18, p = 0.03), CD244 (OR: 1.07, 95% CI: 1.00-1.13, p = 0.036), and IL33 (OR: 1.08, 95% CI: 1.00 -1.17, p = 0.048) were identified as proteins associated with elevated AD risk. Conversely, three inflammatory proteins exhibited a protective effect against AD, including NRTN (OR: 0.91, 95% CI: 0.85-0.99; p = 0.019), CCL4 (OR: 0.95, 95% CI: 0.91-1.00, p = 0.029), and MMP1 (OR: 0.93, 95% CI: 0.87-1.00, p = 0.049). Notably, according to the gene-drug analysis, TSLP, S100A12, CD244, CCL4, and MMP1 were identified as druggable. Additionally, MMP1 (PP4 = 0.92) and CCL4 (PP4 = 0.87) in the prefrontal cortex had the strongest colocalization evidence (PP4 > 0.85), suggesting they could potentially serve as novel therapeutic targets for AD. Integrative genetic analyses indicate that genetically determined circulating levels of TSLP, S100A12, CD244, CCL4, and MMP1 exert causal effects on AD risk. These findings nominate all five proteins as potential therapeutic targets, with MMP1 and CCL4 representing priority candidates warranting further mechanistic investigation and clinical validation.
An et al. (Fri,) studied this question.