Blood metabolites reflect the effect of gut microbiota on differentiated thyroid cancer: a Mendelian randomization analysis

Abstract

Background: Studies have linked gut microbiome and differentiated thyroid cancer (DTC). However, their causal relationships and potential mediating factors have not been well defined. Our study investigated the causal relationships between the gut microbiome, papillary thyroid cancer (PTC) and follicular thyroid cancer (FTC), as well as the mediating effect of potential blood metabolites, using genetic approaches.

Methods: Leveraging the summary statistics of gut microbial taxa, blood metabolites, PTC and FTC from the largest genome-wide association studies (GWAS) to date, we applied the bidirectional and mediation Mendelian randomization (MR) design. The multivariable MR approach based on Bayesian model averaging (MR-BMA) was used to prioritize the most likely causal taxa. Furthermore, metabolic pathway analysis was performed via the web-based Metaconflict 4.0.

Results: After sensitivity analyses, we identified 4 taxa, 19 blood metabolites, and 5 gut bacterial pathways were causally associated with PTC. Similarly, 3 taxa, 31 blood metabolites, and 3 gut bacterial pathways were found to be causally associated with FTC, with 2 blood metabolites exhibiting bidirectional causal relationships. Metabolic pathway analysis revealed 8 significant pathways in PTC and FTC. MR-BMA analysis pinpointed species Bifidobacterium longum as the primary causal taxon for PTC and genus Bacteroides for FTC. The mediation MR analysis showed that sphingomyelin (d18:2/23:0, d18:1/23:1, d17:1/24:1) and 2-hydroxysebacate mediated the causal effects of specific gut microbiota on PTC and FTC, respectively.

Conclusion: The study suggested a causal relationship between several gut microbial taxa and DTC, and that specific blood metabolites might mediate this relationship.

Keywords: Blood metabolites; Follicular thyroid cancer; Gut Microbiota; Mendelian randomization; Papillary thyroid cancer.

Fig. 1

Study design and flowchart. Footnote: MR, Mendelian randomization; DTC, Differentiated thyroid cancer

Fig. 2

A The classification of microbial taxa included in the current study. Each dot represents a microbial taxon. The main branches represent different microbial phyla, and each phylum is shown in a different specific color. Within each main branch (the same phylum), there are additional branches representing further classifications as class, order, family, genus and species, from the inside to the outside. The suggestively significant taxa found in this study were labelled. B-H The forest plot shows the causal estimates between microbial taxa and papillary thyroid cancer or follicular thyroid cancer. The prefixes o., f., g. and s. in the taxa column indicated order, family, genus and species respectively. Footnote: OR, odds ratio; CI, confidence interval

Fig. 3

A Forest plot of the causal associations between blood metabolites and papillary thyroid cancer. B Forest plot of the causal associations between blood metabolites and follicular thyroid cancer. Footnote: Dots depict the point estimate. Horizontal bars depict 95% confidence interval (CI); RAPS, robust adjusted profile score; CML-MA, constrained maximum likelihood and model averaging

Fig. 4

A Enriched significant metabolic pathways of papillary thyroid cancer and follicular thyroid cancer. B Volcano plot showing the causal estimates of 205 microbiota metabolism pathways on papillary thyroid cancer with inverse–variance weighted method. C Volcano plot showing the causal estimates of 205 microbiota metabolism pathways on follicular thyroid cancer with inverse–variance weighted method

Fig. 5

The specific blood metabolites mediated the causal effect of microbial taxa on papillary thyroid cancer or follicular thyroid cancer. Footnote: The β value between microbial taxa and metabolites and cancer are the MR estimates using the inverse–variance weighted method. Mediation effect was derived by using the delta method