Gut Microbiota May Play a Significant Role in the Pathogenesis of Graves' Disease

Abstract

Background: Gut microbiota are considered to be intrinsic regulators of thyroid autoimmunity. We designed a cross-sectional study to examine the makeup and metabolic function of microbiota in Graves' disease (GD) patients, with the ultimate aim of offering new perspectives on the diagnosis and treatment of GD. Methods: The 16S ribosomal RNA (rRNA) V3-V4 DNA regions of microbiota were obtained from fecal samples collected from 45 GD patients and 59 controls. Microbial differences between the two groups were subsequently analyzed based on high-throughput sequencing. Results: Compared with controls, GD patients had reduced alpha diversity (p < 0.05). At the phylum level, GD patients had a significantly lower proportion of Firmicutes (p = 0.008) and a significantly higher proportion of Bacteroidetes (p = 0.002) compared with the controls. At the genus level, GD patients had greater numbers of Bacteroides and Lactobacillus, although fewer Blautia, [Eubacterium]_hallii_group, Anaerostipes, Collinsella, Dorea, unclassified_f_Peptostreptococcaceae, and [Ruminococcus]_torques_group than controls (all p < 0.05). Subgroup analysis of GD patients revealed that Lactobacillus may play a key role in the pathogenesis of autoimmune thyroid diseases. Nine distinct genera showed significant correlations with certain thyroid function tests. Functional prediction revealed that Blautia may be an important microbe in certain metabolic pathways that occur in the hyperthyroid state. In addition, linear discriminant analysis (LDA) and effect size (LEfSe) analysis showed that there were significant differences in the levels of 18 genera between GD patients and controls (LDA >3.0, all p < 0.05). A diagnostic model using the top nine genera had an area under the curve of 0.8109 [confidence interval: 0.7274-0.8945]. Conclusions: Intestinal microbiota are different in GD patients. The microbiota we identified offer an alternative noninvasive diagnostic methodology for GD. Microbiota may also play a role in thyroid autoimmunity, and future research is needed to further elucidate the role.

Keywords: 16S rRNA sequencing; Graves' disease; alpha diversity; gut microbiota; metabolism.

FIG. 1.

General sequencing characteristics of the 105 fecal samples and comparison of alpha diversity in GD patients and HC. (A) Rank-abundance curves of the gut microbiota at the OTU level. (B) Rarefaction curves (SOBs index) of the gut microbiota at the OTU level. (C) Comparison of alpha diversity in GD patients and HC at the OTU level. The SOBs, CHAO, and ACE indices reflect the abundance of microbiota. The SHANNON and SIMPSON indices reflect the alpha diversity, and the COVERAGE index reflects the sequence coverage. **p < 0.01; ***p < 0.001. GD, Graves' disease; HC, healthy controls; OTU, operational taxonomic unit.

FIG. 2.

Altered composition of gut microbiota in GD patients compared with HC. (A) ANOSIM showed that the difference between the GD patients and the HC was significant (p = 0.001) at the OTU level. (B) Analysis of beta diversity using PLS-DA revealed that the microbial composition found in GD patients clearly differed from that found in HC. One dot in the figure represents one sample. (C) The species Venn diagram at the phylum level. Overlapping parts in the figure indicate common species. (D) The species Venn diagram at the genus level. (E) Composition of the gut microbiota at the phylum level. (F) Composition of the gut microbiota at the genus level. ANOSIM, analysis of similarities; PLS-DA, partial least squares discriminant analysis.

FIG. 3.

Statistical analysis of taxa across GD patients and HC. (A, B) Comparison between the two groups at the phylum and genus levels. *p < 0.05; **p < 0.01; ***p < 0.001. (C, D) The LEfSe was used to identify the species that significantly differed between GD patients and HC at the phylum and genus levels. Only taxa meeting a significant LDA threshold value of >3 and p < 0.05 are shown. (E) Comparison across GD subgroups (GD with and without Hashimoto's thyroiditis). (F) Correlation heat-map analysis between genera and thyroid function tests. Red represents a positive correlation, and green represents a negative correlation. LDA, linear discriminant analysis; LEfSe, linear discriminant analysis effect size.

FIG. 4.

The relationship between the predictive genera identified by LEfSe and thyroid function tests as well as metabolic functions. (A) Pearson correlation analysis was used to determine the relationships between genera and thyroid function tests. Red represents a positive correlation, and green represents a negative correlation. *p < 0.05, **p < 0.01. (B) Pearson correlation analysis was used to determine the relationships between genera and metabolic functions. Red represents a positive correlation, and green represents a negative correlation. *p < 0.05; **p < 0.01; ***p < 0.001. (C) The ROC curve was used to assess the diagnostic accuracy of the top nine genera based on LEfSe results. (D) Species cluster analysis confirmed that the two groups could be distinguished by these nine specific microbiota. fT3, free triiodothyronine; fT4, free thyroxine; ROC, receiver operating characteristic; TG, thyroglobulin; TGAB, thyroglobulin antibody; TMAB, thyroid microsomal antibody; TPOAB, thyroid peroxidase antibody; TRAB, thyroid stimulating hormone receptor antibody; TSH, thyrotropin; TT3, total triiodothyronine; TT4, total thyroxine.