Causal effect of gut microbiota and diabetic nephropathy: a Mendelian randomization study

Abstract

Background The interaction of dysbiosis of gut microbiota (GM) with diabetic nephropathy (DN) drew our attention and a better understanding of GM on DN might provide potential therapeutic approaches. However, the exact causal effect of GM on DN remains unknown. Methods We applied two-sample Mendelian Randomization (MR) analysis, including inverse variance weighted (IVW), MR-Egger methods, etc., to screen the significant bacterial taxa based on the GWAS data. Sensitivity analysis was conducted to assess the robustness of MR results. To identify the most critical factor on DN, Mendelian randomization-Bayesian model averaging (MR-BMA) method was utilized. Then, whether the reverse causality existed was verified by reverse MR analysis. Finally, transcriptome MR analysis was performed to investigate the possible mechanism of GM on DN. Results At locus-wide significance levels, the results of IVW suggested that order Bacteroidales (odds ratio (OR) = 1.412, 95% confidence interval (CI): 1.025–1.945, P = 0.035), genus Akkermansia (OR = 1.449, 95% CI: 1.120–1.875, P = 0.005), genus Coprococcus 1 (OR = 1.328, 95% CI: 1.066–1.793, P = 0.015), genus Marvinbryantia (OR = 1.353, 95% CI: 1.037–1.777, P = 0.030) and genus Parasutterella (OR = 1.276, 95% CI: 1.022–1.593, P = 0.032) were risk factors for DN. Reversely, genus Eubacterium ventriosum (OR = 0.756, 95% CI: 0.594–0.963, P = 0.023), genus Ruminococcus gauvreauii (OR = 0.663, 95% CI: 0.506–0.870, P = 0.003) and genus Erysipelotrichaceae (UCG003) (OR = 0.801, 95% CI: 0.644–0.997, P = 0.047) were negatively associated with the risk of DN. Among these taxa, genus Ruminococcus gauvreauii played a crucial role in DN. No significant heterogeneity or pleiotropy in the MR result was found. Mapped genes (FDR < 0.05) related to GM had causal effects on DN, while FCGR2B and VNN2 might be potential therapeutic targets. Conclusions This work provided new evidence for the causal effect of GM on DN occurrence and potential biomarkers for DN. The significant bacterial taxa in our study provided new insights for the ‘gut-kidney’ axis, as well as unconventional prevention and treatment strategies for DN.