Investigating the causal relationship between the gut microbiome and rheumatoid arthritis: mediating effects of immune cells

Abstract

Background: Rheumatoid arthritis (RA) is a complex autoimmune and inflammatory disease that significantly impacts the quality of life for millions worldwide. In recent years, gut microbiota has garnered extensive attention as a potential health-modulating factor, with associations identified between it and various diseases, including RA. This study aims to investigate the causal relationship between gut microbiota and RA using Mendelian Randomization (MR) analysis, and further examines the mediating role of immune cells in this connection.

Method: A MR analytical method was employed by us, integrating genome-wide association study (GWAS) data from FinnGen, MiBioGen, and research led by Valeria Orrù and her team to systematically examine the relationships between gut microbiota, immune cells, and RA. Initially, we performed a bidirectional univariable MR analysis to examine the relationship between gut microbiota and RA, consciously avoiding any possible reverse causal influences. Following this, we applied multivariable MR adjustments on gut microbiota that showed positive associations and employed a two-step methodology to examine the overall genetic predictive role of immune cell-mediated gut microbiota in the risk of developing RA.

Result: Our results demonstrate notable causal connections between different gut microbiota and RA. In particular, Mollicutes, Ruminococcaceae UCG002, and Butyricimonas displayed positive associations with RA, while other microbiota, including Rikenellaceae, Lactobacillaceae, and Veillonella, showed negative associations. Additionally, we identified a reduction in the abundance of certain microbiota, including Lachnospiraceae and Ruminococcus1, which were excluded from our study and validated for analytical accuracy using methods such as "leave-one-out." Immune cells, including CD3 found on activated CD4 regulatory T cells that express CD39, serve a mediating function in the development of RA. To summarize, our research focused on the species Butyricimonas id. 945, recognizing immune cells as crucial contributors to the relationship between genetic predictions of gut microbiota and RA.

Conclusion: This research clarifies the intricate causal links between gut microbiota and RA, emphasizing the crucial mediating function of immune cells in this mechanism. These findings not only enhance our understanding of the pathogenesis of RA but also provide new perspectives and potential intervention targets for future prevention and treatment strategies. Future research should further investigate the specific mechanisms underlying the interactions among gut microbiota, immune cells, and RA, while considering the validation of these findings across diverse populations.

Keywords: Genetic predictions; Gut microbiota; Immune cells mediation; Mendelian randomization; Rheumatoid arthritis.

Fig. 1

Mendelian randomized (MR) design flow chart. The study initiated with a bidirectional univariable MR (UVMR) analysis to explore the association between gut microbiota and rheumatoid arthritis (RA). Following the exclusion of reverse causation, multivariable MR was employed to account for bacteria positively associated with RA. Subsequently, a two-step method was utilized to examine the overall genetic predictive influence of immune cell-associated gut microbiota on RA risk. In the first step, conventional UVMR analyses were conducted for gut microbiota positively correlated with immune cells to determine β1 (p < 0.05). In the second step, multivariable MR (MVMR) was performed using the identified positive immune cells and their associated gut microbiota to calculate β2 (p < 0.05). The total effect observed in the MVMR analysis, labeled as βall, represents the combined influence of gut microbiota on RA. The mediated effect was computed as β1β2, while the direct effect was represented as βall-β1β2. The ratio of the mediated effect was expressed as β1*β2/βall

Fig. 2

Forest map of causal effect of intestinal flora on RA. Mendelian randomisation analysis using inverse variance weighting approach showing significant positive associations between rheumatoid arthritis (RA) and various microbial taxa, including MollicutesRF9.id.11579, unknownfamily.id.1000005471, unknowngenus.id.1000005472, RuminococcaceaeUCG002.id.11360, and Butyricimonas.id.945. Negative associations were observed with Rikenellaceae.id.967, Lactobacillaceae.id.1836, Acidaminococcaceae.id.2166, Veillonella.id.2198, Prevotella9.id.11183, Prevotella7.id.11182, Paraprevotella.id.962, and Sellimonas.id.14369. Odds ratios (OR) and 95% confidence intervals (CI) are provided, with p-values < 0.05 indicating statistical significance

Fig. 3

Forest map of causal effect of RA on intestinal flora. The MR analysis conducted in this study unveiled significant associations between rheumatoid arthritis (RA) and the gut microbiome. Notably, RA was found to be associated with a decreased abundance of several bacterial taxa, including the Lachnospiraceae family (identified by ID 1987), the Ruminococcus1 genus (ID 11373), the Ruminococcaceae UCG009 genus (ID 11366), the Favonifractor genus (ID 2059), and the Erysipelotrichaceae UCG003 genus (ID 11384). Conversely, an increase in the abundance of the Eubacterium fissicatenagroup (ID 14373) was observed in individuals with RA. Due to the clear causal relationships identified between these taxa and RA, they were excluded from subsequent analyses in our study

Fig. 4

Causal effect of positive and positive bacteria on RA (corrected with multivariate). Multivariate Mendelian randomization (MVMR) analysis using the inverse-variance weighted method identified significant negative correlations between family.unknownfamily.id.1000005471 (OR = 0.85, 95% CI: 0.74–0.98) and genus.Butyricimonas.id.945 (OR = 0.85, 95% CI: 0.74–0.98) with rheumatoid arthritis (RA), with p-values < 0.05. Other microbial taxa, including Ruminococcaceae UCG002, Prevotella9, Prevotella7, Paraprevotella, Sellimonas, and Butyricimonas, did not show significant associations with RA (p > 0.05). Heterogeneity, horizontal pleiotropy, and MR-PRESSO analyses confirmed the robustness of these findings

Fig. 5

Causal effect of two positive bacteria on RA funnel plot, “leave-one-out” method plot and scatter plot. A shows a funnel plot indicating no significant publication bias, supporting the reliability of the study results. B illustrates a leave-one-out plot revealing that removing any single study does not alter the overall results, demonstrating the robustness of the conclusions. C displays a scatter plot indicating high precision in the analysis of the relationship between family.unknownfamily.id.1000005471 and RA. For genus.Butyricimonas.id.945, D shows a funnel plot indicating good consistency among studies, despite minor asymmetries. E demonstrates that excluding individual studies does not significantly impact the overall conclusion, reaffirming the robustness of the results. F presents a scatter plot suggesting a clear association between genus.Butyricimonas.id.945 and RA, despite some variability

Fig. 6

Forest map of causal effect of positive intestinal flora on immune cells. Forest plot illustrating the causal relationships between genus Butyricimonas id. 945 and immune cell phenotypes. The Inverse-Variance Weighted (IVW) approach from Mendelian Randomization (MR) was applied to assess these associations. The plot demonstrates a significant negative relationship (p < 0.05) between genus Butyricimonas id. 945 and various immune cell markers. Specifically, the plot shows the odds ratios (OR) and 95% confidence intervals (CI) for changes in CD3 expression on CD39 + activated CD4 regulatory T cells (OR: 0.76, 95% CI: 0.59–0.97), CD28 expression on CD4 regulatory T cells (OR: 0.80, 95% CI: 0.65–1.00), CD3 expression on CD39 + CD4 T cells (OR: 0.77, 95% CI: 0.61–0.98), CD25 expression on CD39 + CD4 regulatory T cells (OR: 0.75, 95% CI: 0.62–0.90), and CD39 levels on monocytes (OR: 0.76, 95% CI: 0.62–0.95). All ORs indicate statistical significance, revealing a noteworthy causal link between genus Butyricimonas id. 945 and the immune cell phenotypes examined