Deciphering genetic causality between plasma BDNF and 91 circulating inflammatory proteins through bidirectional mendelian randomization

Prior studies reported an association between the levels of brain-derived neurotrophic factor (BDNF) circulating in the bloodstream and those of different inflammatory factors. However, their causal relationship remains unclear. Here, we performed a Mendelian randomization (MR) study to investigate the causal relationships between plasma BDNF levels and 91 circulating inflammatory proteins to shed light on the possible role of BDNF in the pathogenesis and progression of inflammation-related neurological diseases in order to distinguish correlation from possible causal effects. Data for plasma BDNF levels were derived from a genome-wide association study (GWAS) encompassing 3,301 European participants. Genetic association estimates for 91 inflammation proteins were extracted from a GWAS meta-analysis that enrolled 14,824 European participants. The primary MR analysis employed the inverse variance weighted (IVW) method and was corroborated by additional methods including MR-Egger, weighted median, weighted mode, and simple mode. Analyses of sensitivity were performed by evaluating the heterogeneity, horizontal pleiotropy, and robustness of the results. Genetic evidence indicated that elevated plasma BDNF levels possibly contribute to decreased concentrations of 13 inflammation proteins (OR: 0.951-0.977), including beta-nerve growth factor (Beta-NGF), caspase 8 (CASP-8), interleukin-15 receptor subunit alpha (IL-15RA), interleukin-17 A (IL-17 A), interleukin-17 C (IL-17 C), interleukin-2 (IL-2), interleukin-20 (IL-20), interleukin-20 receptor subunit alpha (IL-20RA), interleukin-24 (IL-24), interleukin-33 (IL-33), leukemia inhibitory factor (LIF), neurturin (NRTN), as well as neurotrophin-3 (NT-3). The associations between BDNF and IL-33 remained statistically significant after FDR correction (FDR > 0.05). Furthermore, reverse MR analysis showed that C-C motif chemokine 23 (CCL23), CUB domain-containing protein 1 (CDCP1), and NRTN is suggestive for a positive causal effect on BDNF plasma levels (OR: 1.240-1.422). Moreover, 5 proteins are likely to be associated with lower plasma levels of BDNF (OR: 0.742-0.971), including adenosine deaminase (ADA), cystatin D (CST5), interleukin-13 (IL-13), interleukin-17 A (IL-17 A), and vascular endothelial growth factor A (VEGF-A). Genetically determined plasma BDNF levels influence IL-33 and are possibly associated with 12 circulating inflammatory proteins. The data suggest that 8 inflammatory proteins exhibit either negative or protective roles to BDNF levels, respectively. Of these, 5 are negatively associated with BDNF levels, while 3 play protective roles. These findings may offer new theoretical and empirical insights into the pathogenesis and progression of inflammation-related neurological diseases.